Retinoblastoma is a malignant tumor arising from the developing retinal cells and can affect both eyes in one-third of the cases (Dimaras et al., 2015). It is the most common intraocular malignancy in infants and children younger than 5 years and represents about 6% of all cancers in this age group (Broaddus et al., 2009; Singh et al., 2015). Retinoblastoma develops because of biallelic inactivation of the RB1 gene (short arm of chromosome 13) and may affect all body cells (heritable or germline) or a single developing retinal cell (nonheritable or nongermline) (Gallie and Soliman, 2016).
Treatment includes chemotherapy (systemic or regional) and focal therapy (laser, thermotherapy, or cryotherapy) in early cases and enucleation in advanced cases. If untreated, retinoblastoma can metastasize and becomes fatal (Zhang et al., 2012).
Apart from RB1 inactivation, a constellation of epigenetic mutations may have a role in retinoblastoma tumorigenesis (Zhang et al., 2012; Benavente and Dyer, 2015). Several oncogenes and tumor-suppressor genes with histone modifications and abnormal DNA methylation were identified in retinoblastoma (Zhang et al., 2012).
Spleen tyrosine kinase (SYK) is a proto-oncogene that is not related to normal eye development, but was found to be highly expressed in retinoblastoma. SYK is implicated in tumor cell survival (Murphree and Triche, 2012; Zhang et al., 2012). Otherwise, SYK has immunomodulatory functions and is expressed by the hematopoietic system where it was implicated in some hematologic malignancies such as leukemia. Drugs targeting SYK have already found their way to clinical trials for adult leukemia (Brennan et al., 2012) and rheumatoid arthritis among other autoimmune diseases (Zhang et al., 2012). In addition, inhibiting SYK using a small-molecule inhibitor resulted in the apoptosis of retinoblastoma cells both in in-vitro and in-vivo settings (Benavente et al., 2013).
In the current study, we studied the expression of SYK in a cohort of Egyptian retinoblastoma patients. Furthermore, the possible effects of chemotherapy on SYK expression were explored to broaden the understanding of retinoblastoma pathogenesis as well as to probe the possibility of potential interaction between chemotherapy and targeted anti-SYK treatment in retinoblastoma patients.
Material and methods
Study design and sample collection
This is a descriptive study of SYK protein expression in retinoblastoma. Tumors were obtained from enucleated eyes for retinoblastoma at the Ophthalmology Department, Alexandria Faculty of Medicine, during the period from January 2011 to April 2016. Informed consent for specimen study for research purposes was obtained at enucleation. The research ethics committee of Alexandria Faculty of Medicine approved the study. Archival paraffin-embedded material (Property of the Pathology Department, Faculty of Medicine, Alexandria University) was used to study histopathologic high-risk factors and for the assessment of SYK expression.
Hematoxylin and eosin (H&E)-stained sections (5 µm) from archival paraffin blocks of retinoblastoma were reviewed and analyzed for the following: histotype confirmation, assessment of degree of necrosis (score 0, none; score 1, <50% of tumor surface area; and score 2, ≥50% of tumor surface area), appraisal of degree of optic nerve invasion (score 0, none; score 1, prelaminar and superficial; and score 2, postlaminar), and evaluation of choroidal involvement [score 0, none; score 2, focal (<3 mm); and score 3, massive (>3 mm)]. Tumor differentiation was graded as follows: score 1, predominantly showing Flexner–Wintersteiner rosettes+retinal fleurettes, +++; score 2, ++, rosettes; score 3, +, sparse rosettes; and score 4, undifferentiated (no rosettes).
Immunostaining for SYK protein expression
Sections of 4 µm thickness were prepared from retinoblastoma paraffin blocks, deparaffinized in xylene, and rehydrated in descending series of ethyl alcohol. The sections were subjected to heat-induced epitope retrieval in a 700 W microwave in 0.01 mol/l citrate buffer (pH 6.0) for 20 min. Hydrogen peroxide 3% was then applied to block endogenous peroxidase activity. The primary antibody (polyclonal rabbit antihuman SYK, Cat# PA5-28574, at a dilution of 1 : 300) purchased from ThermoFisher Scientific (Waltham, Massachusetts, USA) was applied overnight at 4°C. The UltraVision LP Detection System (Cat# TL-015HD) from ThermoFisher Scientific was used for antigen visualization. The immunohistochemical reactions were developed with diaminobenzidine and the sections were counterstained with Harris hematoxylin. Immunostaining was manually processed, with appropriate positive controls (lymphoma) and negative controls (antibody omission) included in each run.
The expected staining reaction was cytoplasmic with sparing of the normal retina. The immunohistochemical reaction for SYK was semiquantitatively graded as follows: score 1+ (<1/3 tumor cells stained), 2+ (≥1/3 but <2/3 of tumor cells stained), and score 3+ (>2/3 of tumor cells stained).
A review of medical records of the children as regards sex, age at presentation, and treatment was performed.
Forty eyes of 39 retinoblastoma cases were enrolled in the present study. Children’s ages ranged between 4 and 84 months with a mean of 23.65±15.8. Seventeen patients were male (42.5%) and 23 were female (57.5%). There were twenty-two right eyes (55%) and 18 left eyes (45%).
Twenty-seven (67.5%) patients received neoadjuvant systemic chemotherapy at the Pediatric Oncology Unit, Faculty of Medicine, Alexandria University (3-day cycles of VEC repeated every 28 days for 4–6 cycles: vincristine 0.05 mg/kg, etoposide phosphate 5 mg/kg, and carboplatin 18.6 mg/kg), which later failed salvage attempts culminating in enucleation, whereas 13 (32.5%) were treated with upfront surgery.
Nerve invasion was not found in 16 (40%) cases; it was confined to the prelaminar portion in 18 (45%) cases and involved the postlaminar area in six (15%) cases. No cases demonstrated extrascleral spread. Choroid invasion was not detected in 29 (72.5%) cases; it was only focal in eight (20%) cases, whereas it was considered massive in three (7.5%) cases. Only one case showed no necrosis (2.5%), 28 (70%) cases showed score 1 necrosis (in <50% of tumor surface area), and necrosis was marked (≥50%) (i.e. score 2) in 11 (27.5%) cases.
Tumor differentiation was scored as follows: score 1, well differentiated [five (12.5%) cases]; score 2, moderately differentiated [12 (30%) cases]; score 3, poorly differentiated [16 (40%) cases], and score 4, undifferentiated [seven (17.5%) cases].
Patient’s age was not statistically related to degree of differentiation (F=2.17, P=0.11). However, using the analysis of avriance test when differentiation scores (3 and 4) were lumped together, the relation with age became statistically significant (F=3.70, P=0.04) denoting better tumor differentiation with younger age.
Cytoplasmic expression of SYK was observed in 100% of the cases while sparing the native nontumorous retina (Figs 1 and 2). Thirteen (32.5%) cases showed score 1 immunostaining (less than one-third of the cells stained), 10 (25%) cases showed score 2 immunostaining (in more than one-third but less than two-thirds of tumor cells), and 17 (42.5%) cases showed the highest score of expression (score 3: in more than two-thirds of tumor cells).
There was no statistically significant difference between treated and untreated cases with respect to child’s age (P=0.7), sex (P=0.3), nerve invasion (P=0.53), choroidal infiltration (P=0.88), tumor laterality (P=0.56), extent of necrosis (P=0.35), or tumor differentiation (P=0.33). There was no statistically significant difference between treated and untreated cases as regards SYK immune expression (P=0.66).
Degree of SYK expression was not related to child’s age (P=0.10), score of tumor differentiation (χ2=4.74, P=0.58), nerve invasion (χ2=3.91, P=0.41), choroidal infiltration (χ2=3.82, P=0.43), or extent of necrosis (χ2=0.12, P=0.94).
In developed countries, retinoblastoma treatment is a success story. Adjuvant chemotherapy has markedly improved the survival of children at risk for metastasis (Eagle, 2009). Conversely, the disease tends to have poor prognosis in developing countries, owing to late diagnosis (Zhang et al., 2012). Despite a high cure rate in developed countries, there are concerns that chemotherapy can potentially cause side effects, including secondary leukemias, deafness, and neuropathy (Zhang et al., 2012). Developing treatment modalities with fewer adverse effects will offer a practical solution to the problem.
In the present work, SYK protein was expressed in all studied retinoblastoma cases while sparing the normal retina, a notion that gained consensus agreement in the literature (Brennan et al., 2012; Murphree and Triche, 2012; Zhang et al., 2012; Benavente et al., 2013). Furthermore, in a study by Brennan et al., (2012), all metastatic retinoblastoma tumor samples stained positive for SYK, rendering it a promising specific biomarker delineating retinoblastoma as the definite origin of a putative metastatic small blue round cell tumor of unknown primary. Furthermore, epigenetic changes have been successfully used to distinguish metastatic disease from second primary tumors (Racher et al., 2016).
A statistically significant correlation was established between younger patient’s age and better degree of tumor differentiation (F=3.70, P=0.04). This is in agreement with the findings of Madhavan et al. (2008) and Eagle (2009), who reported ‘more numerous Flexner–Wintersteiner rosettes in eyes enucleated from younger infants’.
Given the ubiquity of SYK expression in all retinoblastomas, the statistically nonsignificant relation between SYK and each of optic nerve invasion (χ2=3.91, P=0.41) and choroidal invasion (χ2=3.82, P=0.43) (both known to be pathological high-risk factors) may denote that SYK expression in retinoblastoma is an early/initiating event unrelated to disease progression. Similarly, acknowledging the fact that degree of differentiation is directly proportional to prognosis in retinoblastoma (Khetan et al., 2006; Singh et al., 2015), the nonsignificant relation between SYK expression and tumor differentiation (χ2=4.74, P=0.58) in our work comes as a further evidence to the early action of SYK rather than being a marker of progression.
Neoadjuvant chemotherapy did not affect the degree of tumor differentiation (P=0.33) or extent of necrosis (P=0.35). The situation in retinoblastoma is clearly not analogous to other pediatric tumors (as neuroblastoma) in which response to chemotherapy is translated into differentiation and massive necrosis. However, this seeming paradox can be attributed to the fact that those ‘treated’ tumors have in fact failed salvage therapy (nonresponders), a notion that justified enucleation.
In the current work, neoadjuvant chemotherapy was not observed to affect the degree of expression of SYK (P=0.66). A possible interpretation would be that SYK is not the sole player in retinoblastoma pathogenesis and that other upstream factors might be linked to tumor response. Once more, the notion that ‘treated’ patients are actually nonresponders or else would not have been enucleated might be held responsible for the statistically nonsignificant difference in SYK expression between treated and untreated cases. In such case, SYK assessment in serum or CSF (both before and after therapy) might constitute a prospect worthy of future research. In addition, being unaffected in the course of chemotherapy may justify therapeutically targeting it in refractory cases.
Expression of SYK in all retinoblastoma cases makes it a promising target for therapy. SYK expression was not correlated with any pathological high-risk factor, suggesting its early role in tumorigenesis. There was no significant difference between patients receiving/not receiving neoadjuvant systemic chemotherapy as regards SYK protein expression.
Conflicts of interest
There are no conflicts of interest.
Benavente CA, Dyer MA (2015). Genetics and epigenetics of human retinoblastoma. Annu Rev Pathol 10:547–562.
Benavente CA, McEvoy JD, Finkelstein D, Wei L, Kang G, Wang Y-D, et al (2013). Cross-species genomic and epigenomic landscape of retinoblastoma. Oncotarget 4:844–859.
Brennan RC, Wilson MW, Dyer MA, Bahrami A (2012). SYK expression in metastatic retinoblastoma. J Clin Oncol 30:e13544).
Broaddus E, Topham A, Singh AD (2009). Incidence of retinoblastoma in the USA: 1975–2004. Br J Ophthalmol 93:21–23.
Dimaras H, Corson TW, Cobrinik D, White A, Zhao J, Munier FL, et al (2015). Retinoblastoma. Nat Rev Dis Primers 1:15021.
Eagle RC Jr (2009). High-risk features and tumor differentiation in retinoblastoma a retrospective histopathologic study. Arch Pathol Lab Med 133:1203–1209.
Gallie BL, Soliman SLambert B, Lyons C (2016). Retinoblastoma. Taylor and Hoyt’s paediatricophthalmology and strabismus
. 5th ed. UK: Elsevier, Ltd.
Khetan V, Chan HSL, Wang LGospodarowicz MK, O’Sullivan B, Sobin LH (2006). Retinoblastoma. Prognostic factors in cancer. 3rd ed. Hoboken, NJ: John Wiley & Sons, Inc. 273–277.
Madhavan J, Ganesh A, Roy J, Biswas J, Kumaramanickavel G (2008). The relationship between tumor cell differentiation and age at diagnosis in retinoblastoma. J Pediatr Ophthalmol Strabismus 45:22–25.
Murphree AL, Triche TJ (2012). An epigenomic mechanism in retinoblastoma: the end of the story? Genome Med 4:15.
Racher H, Soliman S, Argiropoulos B, Chan HSL, Gallie BL, Perrier R, et al (2016). Molecular analysis distinguishes metastatic disease from second cancers in patients with retinoblastoma. Cancer Genet 209:359–363.
Singh L, Pushker N, Saini N, Sen S, Sharma A, Bakhshi S, et al (2015). Expression of pro-apoptotic Bax and anti-apoptotic Bcl-2 proteins in human retinoblastoma. Clin and Exp Ophthal 43:259–267.
©2016Egyptian Journal of Pathology
Zhang J, Benavente CA, McEvoy J, Flores-Otero J, Ding L, Chen X, et al (2012). A novel retinoblastoma therapy from genomic and epigenetic analyses. Nature 481:329–334.