Clinicopathological Presentation of Retinoblastoma in Ethiopia

Retinoblastoma (RB) is the most common intraocular malignancy of childhood [1]. It represents approximately 4.0% of all pediatric malignancies [2]. Its occurrence is explained by the “two-hit hypothesis” – a mutation in a single allele of the tumor suppressor gene RB1, either inherited or spontaneous, followed by a “hit” to the second allele leads to tumor initiation [3].

The incidence in various well-studied population groups around the world varies from 1 in 15,000 to 1 in 20,000 live births [4-6]. There is no sexual predilection and it is inherited in an autosomal dominant way. Forty percent of the patients have bilateral (heritable) RB and 60% have unilateral (85% nonheritable; 15% heritable) RB [7]. Of heritable cases, 3.2% develop trilateral RB, which is unilateral or bilateral RB with a midline intracranial tumor [8].

Nearly two-thirds of patients present with leukocoria [9]. Strabismus, secondary glaucoma, proptosis, anterior chamber inflammatory signs, and spontaneous hyphema are the other modes of its presentation [8]. In developing countries, presentation is often late and is characterized by orbital involvement and metastasis [9-12].

Most tumors (>90%) occur before age 2 years and are diagnosed before age 5 years [9]. The median age at diagnosis in the absence of a family history is approximately 24 months for unilateral cases. Those with a family history or bilateral cases typically have an age of diagnosis that is significantly younger, approximately 12 months [13].

Survival of children with RB is highly dependent on the stage of disease of presentation. Hence, understanding the modes of presentation is important for timely diagnosis.

Ethiopia claims 19% of all cases of RB in select sub-Saharan African countries, and it is likely that the number of cases is underreported [14]. This study was conducted to describe the clinicopathological presentation of RB in patients seen at Ethiopia’s major pediatrics eye referral center.

This cross-sectional study was approved by the Institute Ethics Committee (Addis Ababa university 101/17/Oph and Emory University IRB 00098268). All study procedures adhered to the principles outlined in the Declaration of Helsinki for human subject research. Written consent was obtained from all parents/guardians of the study participants.

The study took place at Menelik II Hospital, the major referral Ophthalmology center receiving and providing ocular services for the entire nation in Ethiopia. The majority of RB patients from Ethiopia and neighboring countries are referred to Menelik II Hospital.

All consecutive patients seen at Menelik II Hospital from May 2015 to December 2019, inclusive, were eligible for the study.

Data were collected using a structured data collection form, through medical chart and patient family. The demographic information collected was: age when the presenting sign was noted, age at presentation, sex, region, district, distance of residence from the RB center, age at presentation to our center, first ocular sign noted by parents, laterality, family history of RB, educational level, and monthly income of the parents was asked. Lag time was defined as the interval between the date of first noticing the symptom by the caregivers and the date of treatment at the RB center. Clinical data (International Intraocular Retinoblastoma Classification [IIRC]) and histological pTNM grade were collected directly from the medical chart [15].

The initial evaluation of the children included complete ophthalmological and systemic examination. The ophthalmological evaluation included (a) a dilated-fundus examination using indirect ophthalmoscope, with scleral indentation under general anesthesia (EUA); (b) ultrasonography of the eye; and (c) computed tomography or magnetic resonance imaging of orbit. The stage of the disease was determined using the IIRC-Los Angeles version [16].

In patients where extraocular dissemination was clinically suspected, systemic evaluations were conducted by pediatrics oncologists. This included a full blood count; a lumbar puncture for cerebrospinal fluid cytological analysis; and a bone marrow aspiration looking for tumor cells and evidence of metastases.

Based on the stage of the disease the treatment was planned. For all subjects who underwent either enucleation or exenteration, eye specimen was sent to the department of pathology at Tikur Anbessa Specialized Hospital. The histopathological study included gross and microscopic features of the specimen by the consultant pathologist. The degree and type of cell differentiation, cell morphology, mitoses, and tumor necrosis were noted. The histological diagnosis was based on the extent of RB invasion into optic nerve, choroid, and anterior chamber and staged by the American Joint Commission on Cancer pathological tumor-node-metastasis classification system [16].

Basic descriptive statistical analysis was done using SPSS 21.0 version (SAS Institute, Cary, NC, USA) software. Statistical association between continuous and categorical variables was computed using Fisher’s exact test and Pearson χ² (χ²) test, respectively. Values of p < 0.05 were considered statistically significant.

A total of 251 eyes of 217 patients were included in the study: 183 (84%) patients were unilateral and 34 (16%) were bilateral. For unilateral patients, the tumor involved left eye in 95 (51.9%) and the right eye in 88 (48.1%) of the cases. There was no patient with trilateral RB. The overall mean age at presentation was 32.5 months, ranging from 4 to 144months ± 20.81 months (Table 1). The median age for bilateral cases was 19.6 months versus 35 months for unilateral children (p < 0.001) (Table 2). Majority of subjects with unilateral RB were aged 2 years and above.

The average travel distance of the study subjects was 326.9 km (2–1,900 km) ± 274.6 km. A higher mean age at first symptom and presentation were statistically associated with a longer travel distance more than 200 km (p = 0.03 and p = 0.01, respectively, Table 2). The common first symptoms noted were leukocoria 183 (84.3%), strabismus 13 (5.9%), and combined strabismus/leukocoria in 8 (3.9%). The mean age at first symptom was 23.9 months (1–132 months) ± 18.87 months. The common presenting clinical signs were leukocoria 81 (37.3%), followed by fungating mass 69 (31.8%) and leukocoria with strabismus 25 (11.5%) (Table 3).

The median lag time was 6 (range 0.5–48) months for unilateral patients and 7 (range 1–32) months for bilateral patients. There was no statistical association between the lag time and laterality at presentation (p = 0.752). The median lag time for intraocular RB was 5 (range 0.5–36) months, and for extraocular RB patients, it was 12 (0.5–48) months. Lag time was strongly significantly longer in extraocular RB patients (p < 0.001) and in patients from the rural part of the country (p = 0.03).

There was no investigation done for evidence of RB1 gene mutation. Only 5 patients (2.3%) had family history of RB. The mean age at presentation for patients with family history of RB was 28.8 months, ranging from 10 to 48months ± 11.61 months. The mean age of children with positive family history was statistically significantly younger (p = 0.004).

Of 173 eyes with intraocular disease, the majority of affected eyes were IIRC group E (56.6%), followed by group D (16.7%), group B (12.1%), group C (11.6%), and group A (3.4%) eyes (shown in Fig. 1).

Extraocular disease was clinically evident, with proptosis, and orbital mass in at least one eye at diagnosis was present in 87 of 217 patients (40.1%). Advanced presentation of RB was not statistically associated with laterality (p = 0.100) (Table 2). Eighty-two percent of patients (n = 64) with extraocular RB came from the rural area (p = 0.003). Of cases with regional lymph node involvement (n = 30, 13.8%), 24 (86.7%) of them had unilateral RB. Similarly, out of 50 cases with distant metastasis, 44 (88%) had CNS involvement.

Of the 251 eyes, 114 (45.4%) underwent primary enucleation (87/114, 76.5%) or exenteration (27/114, 23.5%). It is worth noting that primary exenterations (16/27, 59.2%) were standard of care prior to 2017, after which it was replaced by primary chemoreduction followed by enucleation. However, in 11/27 (40.7%) eventual exenteration was performed because the fungating mass proved resistant to the initial chemotherapy.

The remaining eyes (137, 54.6%) underwent salvage therapy of intravenous chemotherapy with focal therapy. At the end of the study period, 46 (33.6%) eyes were salvaged, 35 (25.5%) were still on treatment, and 56 eyes (40.9%) were lost to follow-up (Table 4).

Histopathological studies were performed for those subjects who underwent enucleation and exenteration (n = 114). Sixty-two (54.4%) of the eyes had high-risk pathological features. Table 4 presents the histological findings and pathological tumor-node-metastasis classification of the tumor (Table 4).

This is the first large prospective descriptive study on RB in Ethiopia and one of few such studies in Africa, describing the clinical and histological features of RB. Studies from different parts of the world have shown a wide variation in the clinical presentation. The mean age overall at the age at diagnosis we report in this study (32.6 ± 20.84 months) is comparable to that reported in other developing countries, including Mozambique (32.5 months) and Ghana (36.3 months) [17, 18]. However, it is much higher compared to reported figures from Korea (21.2 months), Turkey (25 months), and USA (18 months) [19-21]. In our study, the mean age of children with unilateral RB was significantly (p < 0.001) higher than the mean age of children with a bilateral, consistent with other investigations [22, 23].

In our series, only 35.5% of patients were diagnosed before 2 years of age, and 6.9% of children were older than 5 years at the time of initial diagnosis, similar to figures from Kenya [24]. All of the patients aged over 5 years had unilateral RB. The mean age at presentation of patients with family history of RB was younger (28.8 months, range 10–48 months) than those without family history. Earlier disease presentation might be the reason for earlier presentation to the RB center, as the lag time is long enough to suggest that there is room for more effective counseling such that offspring of RB survivors are diagnosed much sooner after birth to facilitate optimal treatment outcomes.

Previous studies have observed that RB has no sex predilection [7, 25]. We observed a slight preponderance of male patients (male-to-female ratio of 1.26:1), consistent with reports from Uganda (1.35:1), Uzbekistan (1.3:1), and Nepal (1.5:1) [26-28]. Prior studies have revealed a gender bias on service utilization in the rural part of Ethiopia, which may be manifesting in our cohort [29]. However, the possible reasons for the sex difference noted in our study require further research.

This study showed that majority of the patients (84.3%) presented with unilateral RB, where similar observations were reported from other African countries including Mozambique (82.4%) and Ghana (82.6%) [17, 20]. While similar to what is reported in Africa, our findings differ from North American and European studies which find bilateral cases to amount to 30–40%. Studies from Sub-Saharan African countries have found 11–33% of patients with RB to have bilateral disease [30-35]. The difference might be due to patients being lost to follow-up before the second eye develops in the bilateral RB cases. While there was no case of trilateral RB, this might have been influenced by financial constraints experienced by 23.6% of patients, who were unable to obtain a diagnostic magnetic resonance imaging or computed tomography.

In developed nations, leukocoria is the most common symptom at presentation, followed by strabismus [8, 35]. Likewise, in our study population, leukocoria was the most common presenting symptom (37.3%); however, a large number had late or advanced disease presentations including fungating mass (31.8%) and proptosis (8.3%). This is comparable to other studies done in developing countries including Kenya (38% leukocoria, 22% proptosis), Mali (38.2% leukocoria, 54.5% exophthalmos), and Congo (49% leukocoria, 28% proptosis) [18, 34, 35]. Yet, when parents were asked to report on the first sign they noticed, late signs were rarely observed (Table 3). Compared to high-income countries, children with RB from low-income countries present at an older age with more advanced disease and a higher rate of metastasis [36].

The successful management of RB depends on the ability to detect the disease while it is still intraocular [32]. Extraocular RB is rare in high-income countries and ranges between 2% and 5% [1]. However, in low- and middle-income countries, orbital invasion of RB remains a challenge [37, 38]. In our study, extraocular spread was seen in 40.1% cases, which falls in the range of 26–50% in studies from different parts of the world [19, 39, 40]. Patients with extraocular disease had a significantly longer lag period than patients with intraocular disease (Table 2), suggesting that delay in presentation might be the main reason for extraocular disease. The majority of patients in this study traveled >200 km to arrive at an RB treatment center, and the mean age at first symptom and presentation is prolonged in this patient group, suggesting that distance may be a factor in the late presentation. However, in a comparative study of RB patients from European and African countries, it was observed that, despite traveling on average a shorter distance travel for RB care than their European counterparts (185.7 km vs. 421.8 km, respectively), African patients presented with more advanced disease at presentation [41]. Clearly, distance is not the sole barrier to accessing care in a timely factor, and its impact is likely compounded by additional factors including available travel options, financial support, family structure, or failure to recognize the signs of RB by health care workers [33, 34, 42, 43].

Histological studies were performed in all cases of RB either enucleated or exenterated, revealing that 54.4% of the enucleated eyes had high-risk pathology for metastasis or evidence of extraocular spread. This is comparable to studies from India (54%) and Mozambique (54.9%), but lower than other African series including Uganda (60%) and Nigeria (69.6%) [11, 17, 44, 45]. A cross-sectional global analysis of RB at the time of diagnosis revealed that a drastic increase in the risk of metastasis is associated with higher cT stage [36].

This study reports a large series of clinical-histological presentation of RB among Ethiopian patients attending the main referral center. Our results show that the majority of Ethiopian children with RB have delayed presentation and late-stage disease at diagnosis.

This finding supports the need for improved public awareness and surveillance for the early diagnosis and referral of children with RB to tertiary care centers. The adoption of a nationwide early detection and treatment policy as well as education of medical staff and the general population could help for early detection of the disease in Ethiopia. Further evaluation of the reasons behind delay in presentation, and in particular determinants of health-seeking behavior which might be unique to the Ethiopian context, is necessary for the development of effective screening and referral networks.

The authors wish to acknowledge coworkers and residents at the Department of Ophthalmology of Addis Ababa University. Special appreciation goes to Dr. Bethlehem Sileshi for her outstanding role during data collection.

The study was conducted following the Helsinki declaration and after it was approved by the Institute Ethic Committee Review (Addis Ababa university 101/17/Oph and Emory University IRB 00098268). Written informed consent was obtained from parents or guardians of the participants.

The authors have no conflicts of interest to declare.

There are no funding sources to declare.

Drafting of the manuscript: Sadik Taju Sherief and Daniel Getaneh Mulatu; revision of the manuscript for important intellectual content: Fran Wu, Jacquelyn O’Banion, Sadik Taju Sherief, and Helen Dimaras; conception and design of study: Sadik Taju Sherief, Daniel Getaneh Mulatu, Fran Wu, and Jacquelyn O’Banion; data acquisition, analysis, or interpretation of data: Sadik Taju Sherief, Daniel Getaneh Mulatu, Fran Wu, Jacquelyn O’Banion, and Helen Dimaras; and approval of final manuscript to be published: Sadik Taju Sherief, Daniel Getaneh Mulatu, Fran Wu, Jacquelyn O’Banion, and Helen Dimaras. All authors have read and approved the final manuscript.

All data generated or analyzed during this study are included in this article and are available from the corresponding author (S.T.S.) upon request.