Retinoblastoma is the most common primary intraocular malignancy of childhood. When the disease is diagnosed in its early intraocular stages, as is usually the case in developed countries, the probability of disease-free survival has been greater than 80–90% for decades 1, 2. These encouraging survival figures are not the result of the use of sophisticated treatments since enucleation of the affected eye, a simple surgical procedure, is usually curative in cases of intraocular retinoblastoma. In developed countries, sophisticated conservative treatments are applied in order to preserve eyes with useful vision without affecting survival 3, 4. Although these results and trends are encouraging, the majority of the 8,000 children affected worldwide have not benefited from these trends 5.
In 2010, it is estimated that up to 40% of the children with newly diagnosed retinoblastoma will die of metastatic disease at a young age (Table I) 6. In developing countries, the diagnosis of retinoblastoma is frequently made at later stages of the disease when extraocular dissemination has already occurred; therefore, ocular and patient survival rates are lower in these countries than in developed countries 7. When retinoblastoma extends outside the eye, it is difficult to cure, even with sophisticated and intense treatments 8. Advanced extraocular retinoblastoma correlates with longer lag times from the onset of symptoms to the diagnosis 9, 10. Another problem critically influencing survival in developing countries is that of families opting out of recommended treatment, but it received less attention than late diagnosis in the literature 11. Therefore, the survival of children with retinoblastoma in developing countries may be more influenced by socioeconomic and cultural factors causing delays in diagnosis that lead to extraocular dissemination than by the unavailability of sophisticated treatments as was described for other malignancies like acute leukemias or other solid tumors 9, 10, 12.
Table I. Estimated World Incidence 6, Mortality 7, and Prevalence of Metastatic Disease 7 at Diagnosis and Families Opting Out of Treatment 7 in Retinoblastoma
Economic setting (GNI)
Estimated number of cases/year
Estimated survival rate (%)
Estimated percentage of cases with metastatic disease at diagnosis (%)
Estimated number of cases with metastatic disease at diagnosis (total)/year
Estimated percentage of cases opting out treatment (%)
Estimated cases opting out of treatment (total)/year
LICs, low-income countries; MICs, middle-income countries, L-MIC, lower middle-income countries, U-MIC, upper middle-income countries, HIC, high-income countries, GNI, gross national income in US dollars.
Countries are classified according to the 2008 World Bank Classification.
Retinoblastoma is usually one of the more common malignancies in pediatric oncology services in developing countries 13 and recent evidence from population-based studies showed an increased incidence of retinoblastoma in some countries compared to western countries 1, 14. Data from national and regional registries confirmed an increased incidence of retinoblastoma in some areas of Brazil 15 and Mexico 16, where the incidence of retinoblastoma is as high as three to seven times that of Western Europe 14. Environmental factors such as a decreased intake of fruits and vegetables during pregnancy 17 were implicated, and a putative role of the human papilloma virus in the tumorigenesis of sporadic retinoblastoma in Mexico was suggested 18 but not confirmed in other areas of the world.
OUTCOME OF RETINOBLASTOMA IN DEVELOPING COUNTRIES
When the outcome of patients with retinoblastoma is stratified by country-type using the World Bank's classification scheme (Table I) as an estimate for a country's level of development, there are significant differences in survival according to this variable 7. Some variables related to the general development of society, such as the Human Development Index, are significantly correlated with the outcome of patients with retinoblastoma 7.
Ranging from 20% to 60%, the survival of patients with retinoblastoma is dismal in most low-income countries (LICs) (Table I) 7, 19, 20. In some African or Asian countries, the survival rate is virtually zero, because most patients do not complete therapy or are lost to follow-up 19–21. Pathologic evaluation of enucleated eyes is seldom done, and relatively intensive chemotherapy is not usually available 19–21. Survival in many middle-income countries (MICs) is over 70%, and some upper MICs that have established multidisciplinary teams and adequate treatment facilities have survival rates near 90% (Table I) 22–25.
STRATEGIES FOR EARLIER DIAGNOSIS AND FOR IMPROVING TREATMENT ADHERENCE
Delayed diagnosis accounts for the increased prevalence of advanced retinoblastoma in developing countries and explains why retinoblastoma is diagnosed at a later age in these countries 9, 10, 26, 27. Brazilian investigators were the first to report the importance of lag time (i.e., the interval between the onset of symptoms and diagnosis) in determining outcome 9. In most cases, the reason for this delay in diagnosis is either because the child was not brought in for medical attention or because the initial diagnosis was incorrect 10.
Media campaigns to increase public awareness have been launched in some countries since it was hypothesized that the general population frequently ignores leukocoria because they do not recognize it as a sign of disease, particularly cancer in children. These campaigns were initially launched in Brazil, and subsequently in many other countries targeting the public and local doctors 28. The goal of awareness campaigns is to improve survival by allowing cases presenting with leukocoria to be diagnosed before they become metastatic and therefore incurable. Only the results of the campaign in Honduras have been reported showing a reduction in the number of cases with extraocular disease from 5.3 to 3.2 cases per year and the proportion of patients with extraocular disease was significantly reduced during the campaign 29. However, these results are preliminary and based on a limited number of patients from a single institution in a short period of time. There are no population-based data showing the effectiveness of awareness campaigns, however, they may be an option in LICs or some lower MICs, where metastatic disease at diagnosis is still a problem and most patients die of advanced disease, but their impact needs to be determined by solid data 29. On the other hand, since survival exceeds 80% in most MICs, they are probably not a priority in that setting. In those countries, campaigns could be implemented to improve the preservation of eyes and vision, but experience from developed countries suggests that this is difficult to achieve because most eyes with leukocoria have little chance of being preserved 30. Only screening for retinoblastoma via regular eye examinations under anesthesia could detect tumors at an earlier stage so that conservative treatment may be successfully administered. However, given the low incidence of retinoblastoma, this screening approach is recommended only for the siblings of patients with retinoblastoma 31. However, this is not followed in developing countries where in some settings only 23% of children with familial retinoblastoma were diagnosed by screening 32. Therefore, the only realistic goal in terms of improving eye preservation in developing countries by early diagnosis would be to adequately screen children with familial retinoblastoma.
Because the countries where late stage at diagnosis is prevalent are generally the ones that also have up to 50% of families opting against recommended therapy, efforts directed toward detecting retinoblastoma earlier should be accompanied by efforts to retain families in the healthcare system once the diagnosis is made. Programs aimed at increasing treatment adherence may have an even greater effect on improving survival than early diagnosis campaigns do. In a study in Indonesia, Sitorus et al. 11 identified the factors leading to parental decisions against recommended treatment; these included limited access to ophthalmologists, living in remote areas, lack of health insurance, and financial cost. They reported that rates of these decisions were higher when the disease was in the earlier intraocular stages, probably because of the parents' rejection of enucleation as a treatment of an otherwise healthy child 11. Therefore, as reported by Sitorus et al. 11 45% of the children whose families opt out of recommended treatment have intraocular disease and only 20% of them will survive. If these children were effectively treated, the impact in survival worldwide would be comparable to a reduction by the half of the number of metastatic cases at diagnosis (Table I). This pattern of parents not accepting the enucleation of the affected eye or defaulting from follow-up after a conservative intent is similar in many areas of the developing world. Rates from 18% to 50% have been reported by groups in from Asia, Africa, and Central America 11, 13, 21, 26, 29, 33. Decisions against enucleating the second affected eye after conservative treatment fails account for a survival rate of less than 60% in patients with bilateral retinoblastoma in Taiwan, a relatively affluent country having a national healthcare program that offers subsidized care 34. A recent study in India found that initial enucleation was not accepted in about 7% of the cases, but children undergoing conservative treatment had more problems with adherence to treatment, since about two-thirds of the eyes that were treated with a conservatively did not receive effective treatment or completed follow up 26. In that study, there were no significant differences in compliance according to gender, religion or place of residency 26. The factors influencing lack of treatment adherence may vary in different areas and the role of the extended family appears to be critical in some African countries 35, 36. After leaving the hospital to consult with the extended family, the child may never be brought back or, as was reported in Indonesia, be brought back several months later, after the family has tried all sorts of traditional remedies and massive proptosis has developed; at this point, there is no longer hope for cure 11.
To plan local strategies to increase treatment adherence, it is important to identify the factors involved in rejection of enucleation. There are many examples of recent initiatives for retinoblastoma programs in Jordan, Central America, Saudi Arabia, and Indonesia to improve adherence through different approaches that are based on local cultural factors and resource availability 11, 37–39. Finally, twinning programs with institutions in developed countries and the action of international organizations have facilitated technology transfer and made conservative treatment possible 38. In some settings, these programs have resulted in increased treatment adherence 38.
CHALLENGES OF TREATING RETINOBLASTOMA IN DEVELOPING COUNTRIES
Because of the high prevalence of advanced disease at diagnosis (Fig. 1), the priorities for managing retinoblastoma in developing countries are somewhat inverse to those in developed countries (Table II). These priorities include the following, in order of their potential impact on survival: (1) treatment of overt extraocular disease, (2) treatment of retinoblastoma with high-risk pathology features, and (3) eye-conserving therapy.
Table II. Summary of Retinoblastoma Management in Developing Countries
COG, Children's Oncology Group; INCTR, International Network for Cancer Treatment and Research; LICs, low-income countries; MICs, middle-income countries; SIOP, Société Internationale D'Onocologie Pédiatrique.
The economic settings were defined by the countries' gross national income per capita, as determined by The World Bank (2008).
Disease free-survival: 70–90%
Inequity of care in smaller distant centers
Establishing multidisciplinary treatment groups in large centers with pediatric oncology services if not available
Cooperative national treatment protocols in Brazil, Mexico, and Turkey
Metastatic disease in <10% of patients at diagnosis
Poor screening of familial cases
Developing a specialized eye pathology team for adequate microstaging of extraocular disease
COG international treatment protocols
∼30–40% of patients have locally advanced disease
Developing programs for screening familial cases
Chemoreduction programs in selected institutions in Brazil, Argentina, Belarus, Venezuela, and other nations
Advanced intraocular bilateral disease
Improving eye-conservative therapies
Increasing prevalence of familial cases
Improving access to adequate treatment in countries with high inequity in this area
Disease-free survival: 50–70%
Creating awareness campaigns for early referral if metastatic disease >20%
Early diagnosis campaigns in Central America and India
Metastatic disease in >10% of patients at diagnosis
Developing programs to prevent treatment nonadherence if >5%
Centers of Excellence in India, China, Iran, Jordan, Thailand, Egypt, and other nations
Twinning programs with institutions in high-income countries to improve survival
Twinning program between Indonesia and The Netherlands, St Jude and Central America, St. Jude and Jordan
Developing radiotherapy facilities for eye-conservative therapy of advanced disease
Disease-free survival: <50%
Creating awareness campaigns for early referral
SIOP initiative in Malawi
Metastatic disease in >30% of patients at diagnosis
Developing programs to prevent treatment nonadherence if >5%
INCTR retinoblastoma strategy group for global outreach
Providing palliative care for patients with metastatic disease
Twinning program between Canada and Kenya
Twinning programs with institutions in high-income countries to improve survival
TREATING PATIENTS WITH OVERT EXTRAOCULAR DISEASE
Overt extraocular retinoblastoma, including distant metastatic dissemination, is more frequent in LICs; as many as 50% of newly diagnosed retinoblastoma cases in some nations involve advanced disease. In contrast, in most MICs, metastatic disease is present at diagnosis in less than 10% of the cases 7, 22, 23. Treatment with neoadjuvant chemotherapy, followed by secondary enucleation and adjuvant therapy is the current approach in most centers 13, 40, 41. Encouraging results have been reported for the treatment of retinoblastoma that is disseminated to the orbit or the preauricular lymph nodes has resulted in disease-free survival rates over 70% 41. Different chemotherapy combinations have been used to treat overt extraocular retinoblastoma; most regimens include a platinum derivative; an alkylating agent, such as cyclophosphamide or ifosfamide, etoposide, and anthracyclines 22, 41, 42. A low intensity regimen using these agents led to 56% survival at 30 months in Mali 13. However, better results with disease-free survival greater than 70% can be achieved with more intensive regimens 41.
Results of using conventional chemotherapy and irradiation for distant metastatic disease have been disappointing in developing countries 42, 43. Patients with metastatic disease outside the central nervous system (CNS) may be cured with high-dose chemotherapy and stem cell rescue 8, 44; however, these treatments usually are not available in LICs or most lower MICs. In cases with initial CNS involvement, subsequent relapse in the CNS often occurs after an initial response and ultimately causes death 42.
Even though curative therapy for metastatic disease may not be available in LICs, newly diagnosed retinoblastoma is usually responsive to low-dose chemotherapy or radiotherapy that could be used as palliation for these patients. In settings where potentially curative therapy for metastatic retinoblastoma is not available, adequate palliation of these children is a priority, however, access to adequate palliative therapy is also usually limited.
TREATING PATIENTS WITH HIGH-RISK PATHOLOGY FEATURES
Patients presenting with high-risk pathology features, such as microscopic tumor invasion of the postlaminar optic nerve (i.e., beyond the lamina cribrosa), choroid, or sclera, are at higher risk of extraocular retinoblastoma relapse. However, the relapse rate is different among the different groups. Such cases are more frequent in developing countries, occurring in more than 50% of children in some MICs compared to developed countries where they usually occur in less than 20% of the enucleated eyes. Studies in many upper MICs have yielded encouraging results, with survival rates greater than 80% for patients with these conditions 23, 45, 46. However, as opposed to other childhood malignancies, there are no randomized studies to support the treatment decisions in these cases. Most of these results were generated in large referral centers, where multidisciplinary groups that treat retinoblastoma have been established but national cooperative groups have been recently established in Brazil, Mexico, Central America, and Turkey in an attempt to generalize these successful results 24, 25, 38. However, until recently, a consensus staging system, common definitions for invasion of the ocular coats and pathology guidelines for the evaluation of enucleated eyes were not available 47, 48. The agreement of such guidelines was a major step toward harmonizing the diagnostic and staging procedures used by different centers in developing countries that would make possible the comparison of treatment results and the design of future prospective studies.
Thus, there are many controversies in the management of these patients such as
(1)The use of adjuvant chemotherapy for patients with high-risk pathology features with low relapse rates. Isolated choroidal invasion is the typical example where a higher risk of extraocular relapse has been reported and adjuvant therapy has been recommended by some authors while others propose withholding adjuvant therapy 49. According to the results of studies conducted in MICs, the relapse rate is near 3% 45; therefore, the role of adjuvant chemotherapy would be difficult to estimate, even with adequately designed prospective studies. However, a prerequisite for safely withholding adjuvant therapy is a reliable histopathological staging of the enucleated eye. This evaluation may not be available in many centers in developing countries, so risk estimation may be misleading. If the eye is not meticulously evaluated and concomitant invasion to other critical ocular coats such as the postlaminar optic nerve and the sclera cannot be adequately ruled out, withholding adjuvant chemotherapy may be unsafe. The availability of treatment of extraocular relapse should be also considered for the decision of using adjuvant therapy in these patients, since once extraocular relapse occurs, it is impossible to retrieve these patients with conventional therapy. Therefore, eye pathology examination done by experienced ocular pathologists is critical for identifying high-risk patients. When it is not available, patients with high-risk disease may be missed and if they are not given adjuvant therapy, they may eventually present an extraocular relapse and if intensive treatment for relapsed extraocular disease is not available, they may die of metastatic retinoblastoma. In these situations, lower dose adjuvant therapy may prevent the occurrence of extraocular relapse in some of these patients with undiagnosed higher risk disease. Therefore, adequate eye pathology services are a priority in most MICs where these patients are common.
(2)Adjuvant therapy of patients with high-risk pathology features with high relapse rate. This group includes children with postlaminar optic nerve with or without negative margins and scleral invasion. In these cases adjuvant chemotherapy is universally recommended. However, the most effective regimen is not known. There are reports suggesting that more intensive regimens are more effective in preventing extraocular relapse 41, 45, 50 and the Children Oncology Group is performing an international study using an intensive regimen for patents with tumor invasion at the resection margin of the optic nerve. This is a common problem in developing countries that can be a consequence of advanced disease, when the tumor extends through the whole optic nerve length; but it also occurs when patients undergo enucleation in nonspecialized centers and an adequate optic nerve stump is not obtained during enucleation, thereby resulting in subsequent neoplastic invasion 51. Orbital irradiation is usually needed in these cases and some groups also advocate intrathecal chemotherapy, but it appears to be unnecessary when intensive systemic chemotherapy is used 51. The use of high intensity regimens like the ones discussed requires adequate supportive care that may not be available in many centers with limited resources. In these cases, the reduction of extraocular relapse may be offset by an increased toxic mortality, so lower dose regimens may be a reasonable option.
Chemoreduction with carboplatin-based regimens followed by focal therapy for eye-conservative treatment is the standard of care in developed countries 3, 4. Results of this treatment in eyes with less-advanced disease are comparable to those of treatment with external-beam radiation therapy (EBRT) but avoid EBRT's late sequelae such as the occurrence of radiation-induced secondary malignancies 52, long-term ocular toxicity, and facial deformity 53. However, the results of chemoreduction and focal therapy in children with more advanced tumors, especially when vitreous seeding is present, are less encouraging, and it has not shown an obvious superiority to EBRT 54. Consequently, enucleation or EBRT are needed in up to two-thirds of the advanced cases 3, 4, 55. On the basis of these results from the USA and Europe, many upper MICs have also implemented programs for chemoreduction and focal therapy with satisfactory results 56–59. However, these programs require a large investment in personnel training and technology to procure sophisticated local treatment modalities 60 which are not always available in some developing countries or they are only available in highly specialized centers in big cities. According to the results in developed countries, the use of chemoreduction and local therapy is justified because it allows the patient to avoid fatal radiation-induced secondary malignancies that develop in as many as 11% of irradiated patients after 30 years 61, but other factors should also be considered to support using chemotherapy instead of EBRT in developing countries. In this setting, toxic mortality was reported after the use of chemoreduction using more prolonged or intensive treatments including the use of drugs with greater long-term toxicity such as cisplatin or cyclophosphamide 37, 56, 59, 62, 63. Problems in adherence to treatment of children receiving chemoreduction were also reported. Only one-third of the patients completed effective treatment in one series 26 and a variable proportion, reaching 20% in one study 59, of children died of extraocular dissemination after their parents decided against enucleation after failure of conservative therapy 56, 59, 63, 64. Children treated with chemoreduction and who have no local therapy will almost invariably experience relapse 65, so if they become lost to follow-up or if their families do not choose enucleation when local therapy fails or is not given, these patients may suffer fatal extraocular dissemination 11.
Therefore, to justify the use of chemoreduction instead of EBRT as an eye-conservative therapy in developing countries, it is important to estimate the number of patients who die of retinoblastoma or toxicity and those who die of radiation-induced secondary malignancies. The precise numbers of this equation vary in different settings, but unfortunately they are difficult to estimate with the information available because little data of secondary malignancies have been published 66, and tumor registries in developing countries are incomplete or nonexistent. Since the occurrence of radiation-induced secondary malignancies may be lower in children older than 1 year at the time of irradiation 67, children from developing countries might have a lower risk of radiation-induced secondary malignancy since they are diagnosed usually after the first birthday 26, 63. However, the access to high quality radiotherapy in developing countries is also limited 68, so the availability of treatment resources also influences the decision making in this setting. Therefore, the strategy for eye-conserving therapy should be adopted considering the clinical and socioeconomical situation as well as the available resources according to each of the socioeconomical settings (Table III).
Table III. Schematic Overview of the Situation in Eye-Conserving Treatment in Developing Countries According to Country Classification, as Determined by The World Bank (2008)
Few patients present with intraocular disease and those who do are usually not candidates to conservative therapy
Patients with intraocular disease may die of extraocular dissemination because of their families opt out of enucleation
Actions to improve acceptance of enucleation as a life saving procedure
Conservative treatments are not a priority
Advanced intraocular disease presenting in children older than 1 year
Increased risk of extraocular dissemination related to poor compliance
Actions to improve compliance to treatment
Low availability of conservative treatments
Increased toxic mortality related to the use of chemoreduction
Development of adequate facilities for external beam radiotherapy
Poor compliance to treatment
Increased need of EBRT because of advanced disease potentially leading to increased risk of second malignancies
Implementation of lower dose chemoreduction regimens
Actions to improve supportive care
Advanced intraocular disease presenting in children older than 1 year
Slight increased mortality related to the use of high dose chemoreduction
Development of centers of excellence for the conservative treatment of retinoblastoma
Availability of conservative treatments in large centers
Lower availability of sophisticated local therapies (iodine plaques, lasers)
Improved access to local treatments to avoid EBRT
Adequate compliance to treatment
Increased need of EBRT because of advanced disease or lack of availability of local therapies potentially leading to increased risk of second malignancies
Actions to improve supportive care
Information on secondary malignancies to plan future strategies
CONCLUSIONS AND ACTIONS NEEDED
We conclude that the outcome of patients with retinoblastoma in developing countries is grossly influenced by the lack of prompt diagnosis, which is determined by socioeconomic and cultural factors that also influence treatment adherence. Socioeconomical development of the society and adequate access to treatment would improve the survival of children with retinoblastoma in these countries. Although better treatment of loco-regional extraocular disease improved the outcome of patients in upper MICs, the ultimate goal of those nations is to diagnose the disease before it is disseminated outside the eye and can still be cured with less-aggressive treatments. Finally, retinoblastoma programs in countries worldwide should consider the social, economic, religious, and financial milieu; doing so might improve communication with patients' families and allow each country to maximize its medical resources.
Dr. Guillermo Chantada was supported by a grant from the Fund for Ophthalmic Knowledge (New York, NY, USA) and the Fundacion Natalie Dafne Flexer de Ayuda al Niño con Cancer (Buenos Aires, Argentina). Dr. Ibrahim Qaddoumi was supported by the American Lebanese Syrian Associated Charities (ALSAC) (Memphis, TN, USA). These organizations had no role in designing or conducting the studies reported in this article. We also thank Angela Mc Arthur for editorial review.