Retinoblastoma: A review of the molecular basis of tumor development and its clinical correlation in shaping future targeted treatment strategies

Abstract

Retinoblastoma is a retinal cancer that affects children and is the most prevalent intraocular tumor worldwide. Despite tremendous breakthroughs in our understanding of the fundamental mechanisms that regulate progression of retinoblastoma, the development of targeted therapeutics for retinoblastoma has lagged. Our review highlights the current developments in the genetic, epigenetic, transcriptomic, and proteomic landscapes of retinoblastoma. We also discuss their clinical relevance and potential implications for future therapeutic development, with the aim to create a frontline multimodal therapy for retinoblastoma.

Retinoblastoma is the most common pediatric eye cancer in children, accounting for about 2.5%–4% of all pediatric cancers in developed countries, with a relatively higher incidence rate of 4.4 per million in moderate to low–income countries in the ages between 1 and 4 years.[1] If left untreated, the condition is lethal, with a global survival rate of less than 30%.[2] The root cause of retinoblastoma is the biallelic inactivation of the RB1 gene, but additional genomic alterations such as genomic gain of 1q, 2p, and 6p, along with the genomic loss of 16q may also contribute to the progression of the tumor.[3]

Available therapeutic options to treat retinoblastoma include enucleation, cryotherapy, transpupillary thermotherapy, plaque brachytherapy, external beam radiotherapy, and systemic chemotherapy. Systemic chemotherapy involves the use of drugs such as carboplatin, cisplatin, vincristine, etoposide, topotecan, cyclophosphamide, and doxorubicin.[4] While the standard treatment of retinoblastoma has been effective in terms of tumor control, its non-selective nature and associated systemic adverse effects pose a major challenge. Recently, super-selective intra-arterial chemotherapy has replaced systemic chemotherapy for the treatment of intraocular retinoblastoma; however, it has ocular and systemic side effects [Table 1].[5] The management of retinoblastoma varies depending on the stage of the disease, the patient’s age at the time of presentation, the tumor’s location, size, and laterality, societal perception, and cost-effectiveness of the treatment.[6] While these methods have proven to be effective for most patients, they have several limitations, including side effects such as partial or complete vision loss, immune response–generated inflammation, dry eye disease, retinal detachment, focal iris atrophy, focal lens opacity, retinal traction, and late ocular and systemic side effects, such as bone marrow suppression, autotoxicity, nephrotoxicity, and high chances of developing cataract, ptosis, and secondary malignancies after treatment.[7] As a result of non-targeted drug deliveries, these therapies can lead to severe complexities that can cause on-target off-tumor toxicity.[8]

Table 1.

Current treatment of retinoblastoma based on the tumor staging with its advantages and limitations

Tumor Type	Stage (International Classification of Retinoblastoma)	Treatment	Advantage	Limitation
Intraocular tumor, Group A-C	0–1	Focal therapy	• Effective against small tumors, with minimal complications	• Non-effective against large tumors
		Brachytherapy	• Suitable for localized retinal tumors not responding to chemotherapy or other modalities	• Retinal scar and associated radiation retinopathy complications
		Chemoreduction	• Satisfactory disease control with tumor regression	• Systemic side effects that are short and long term • Risk of tumor recurrence
Intraocular tumor, Group D-E	0–1	Chemoreduction	• Tumor regression especially for bilateral cases	• Systemic side effects • Tumor recurrence/metastasis
		Periocular chemotherapy	• Effective in treatment of vitreous seeds	• Periorbital erythema/edema/orbital fibrosis and associated complications
		Intravitreal chemotherapy	• Highly effective in vitreous seed treatment	• Intraocular toxicity of chemotherapeutic agents, injection-related complications, remote risk of tumor dissemination and metastasis
		Intra-arterial chemotherapy	• Effective against advanced intraocular retinoblastoma, can help eye and vision salvage in advanced tumors	• High cost, learning curve, procedure-related side effects, ocular and systemic side effects, chemotherapeutic agents related side effects
		Enucleation	• Allows histopathological evaluation of tumor • Effective tumor control of advanced intraocular tumors	• Cosmetic deformity • No organ salvage • Poor treatment acceptance
High-risk Retinoblastoma	2	Orbital external beam radiotherapy	• Minimizes chances of recurrence and metastasis	• Radiation-related side effects like orbital growth retardation, risk of second malignancy in germline tumors
		Adjuvant chemotherapy	• Reduces the risk of systemic metastasis	• Systemic toxicity related to chemotherapeutic agents
Extraocular Retinoblastoma	3A/B	High-dose chemotherapy	• Reduces risk of metastasis and recurrence	• Vision loss • Autotoxicity, nephrotoxicity • Bone marrow suppression • Other systemic side effects
		Extended enucleation/exenteration	• Option for management of massive orbital invasion in disorganized eyes without prospect of vision	• Cosmetic deformity
		External beam radiation therapy (EBRT)	• Option for advanced disease as adjuvant treatment	• Conjunctival xerosis, contracted socket • Risk of second malignancy in the field of radiation in germline cases
Tumor with CNS metastasis	4	Systemic high dose chemotherapy	• Prolongs survival	• Drug-induced toxicity • Poor prognosis • Palliative treatment
		Intrathecal chemotherapy	• Prolongs survival	• Drug-induced toxicity
		Enucleation/exenteration	• Option for management of massive orbital invasion in disorganized eyes without prospect of vision	• Cosmetic deformity • Orbital bony growth retardation

Therefore, a paradigm shift from conventional treatment options to targeted therapies that offer a stronger anti-tumor response with minimal complication is needed. Such futuristic therapies have minimal complications and offer better patient outcomes. Advanced techniques such as high-throughput screening (HTS), next-generation sequencing (NGS), and gene expression analysis, in addition to RB1 inactivation, have identified various genetic targets that play crucial roles in retinoblastoma progression. Such novel biomarkers can help drive the patient’s treatment toward a more precise and accurate therapy resulting in a better prognosis. We present here a comprehensive review of the advancement made in the genetic, epigenetic, transcriptomic, and proteomic aspects of retinoblastoma over the past 21 years, which will pave way for development of targeted therapy for retinoblastoma.

A Timeline of Events in Understanding the Molecular Circuitry Involved in the Development of Retinoblastoma

In 1971, Alfred Knudson proposed the two-hit hypothesis that suggested the necessity of two mutational hits for the progression of retinoblastoma tumors.[9] The timeline of discovery of the genes involved in retinoblastoma, apart from RB1, that were identified with the help of multi-omics studies is summarized in Fig. 1.

Figure 1.

Genetic timeline of retinoblastoma tumorigenesis (1971,[9] 1978,[10] 1984,[11] 1986,[12] 2001,[13] 2005,[14] 2007,[15] 2008,[16] 2014,[17] 2016,[18] 2020,[19] 2021[20])

Various genetic, transcriptomic, and epigenetic events that play important roles in the growth, development, and progression of retinoblastoma have been discovered. All of these molecular findings have been made possible with the help of omics studies as follows:

Potential target gene identification from genomics and transcriptomics profiling of retinoblastoma

Besides the genomic alterations in RB1 and MYCN, various omics technologies, including microarray, whole-genome sequencing (WGS), and RNA-sequencing have been utilized in recent years to identify several “driver genes” in both localized and metastatic retinoblastoma.[21] The correlation between these gene expression patterns and the tumor phenotype is crucial for clinical purposes as it clarifies the significance of these expression levels and how they can aid in the diagnosis, prognosis, and treatment approach for patients. Through comparative genome hybridization (CGH) analysis, KIF14 was found in the regions of 1q gain in retinoblastoma tumors.[22] Madhavan et al.[23] validated the overexpression of KIF14 and E2F3—both of which are associated with growth and proliferation of tumors—in a large set of retinoblastoma tumor samples using quantitative PCR (qPCR). Overexpression of MDM2, validated via qPCR, is known to inhibit the p53 apoptotic pathway that allows tumor progression in retinoblastoma.[24] Nutlin 3, a small molecule inhibitor that disrupts the interaction between MDM2 and p53, has shown promising results in vitro. However, for optimal efficacy, Nutlin 3 was administered in combination with topotecan via subconjunctival injection in a preclinical study, resulting in a significant reduction in tumor volume. Furthermore, patients treated with this combination therapy showed significantly improved survival rates due to increased tumor necrosis and activation of the p53 pathway, compared to those treated with multimodal chemotherapeutic regimens like vincristine, etoposide, carboplatin, or topotecan.[25] Using single-cell RNA sequencing, the expression levels of UBE2C—an oncogene that promotes growth and proliferation of cancer cells—was estimated for retinoblastoma.[26] It was observed that UBE2C levels were remarkably higher in retinoblastoma tissues of metastatic patients younger than 3 years of age, suggesting that the gene is activated in the malignant stage. In vivo knockdown of UBE2C showed that the tumor cell growth significantly decreased in all UBE2C–knocked down cells compared to the normal ones, forming comparatively smaller colonies. Survival analysis demonstrated that the patients with higher UBE2C expression levels had a significantly lower survival rate, suggesting that it could be considered as a potential prognostic biomarker. Therefore, this underscores the clinical importance of UBE2C in retinoblastoma tumorigenesis.[26] OTX2 gene is crucial for early neurogenesis but can act as an oncogene if expressed in later stages of life.[27] Its overexpression is linked to the progression of retinoblastoma, as confirmed by qPCR analysis. Pharmacological inhibition of OTX2 using all-trans retinoic acid induces apoptosis, decreases cell proliferation, and reduces colony formation. Furthermore, a combination treatment of vincristine and all-trans retinoic acid resulted in a significant increase in the inhibitory effect on retinoblastoma cells.[27] NEK7, which is crucial for cell division, is upregulated in retinoblastoma tumors, as confirmed via qPCR analysis.[28] Knockdown of NEK7 using Lentivirus-mediated RNA interference resulted in significant inhibition of cell growth, cell cycle arrest, and impaired colony formation ability. These findings suggest that NEK7 could be a potential therapeutic target for treating retinoblastoma.[28] SKP2—known to downregulate p27 and inhibit the surveillance apoptotic pathway—is found to be overexpressed in retinoblastoma, as confirmed via qPCR analysis.[29] In a preclinical study, Aubry et al.[30] demonstrated the potential of a novel small molecule NAE-inhibitor MLN4924 in targeting SKP2 for retinoblastoma. The study found that both RB1-null and MYCN-amplified retinoblastomas responded well to the drug, causing successful cell cycle arrest, apoptosis, and SKP2 inhibition in cancer cells. These findings highlight the promising potential of MLN4924 as a targeted therapeutics approach for treating retinoblastoma.[30]

The loss of CDH11 expression, a cadherin involved in cell adhesion, was observed in retinoblastoma cells that were invading the optic nerve and choroid via CGH analysis and as confirmed by qPCR analysis.[31] This loss of CDH11 was also found to be associated with diffuse vitreous seeding, indicating that changes in cadherin-mediated cell adhesion mechanism may contribute to the invasive properties of retinoblastoma tumor cells.[31] The transcription regulatory factor BCOR was found to be downregulated in retinoblastoma, as identified via NGS MSK-IMPACT analysis.[20] BCOR, which is commonly mutated in various malignancies, plays a crucial role in the normal development of the retina and sclera of the eye. Its downregulation is considered to be the most frequent non-RB1 genome abnormality in retinoblastoma and is associated with poor prognosis and worsened metastasis-free survival.[20] RNA-sequencing analysis reveal a decrease in ARHGAP9 expression levels.[32] The downregulation of ARHGAP9 was found to significantly affect cancer cell susceptibility toward chemotherapeutic drugs, such as carboplatin and etoposide, leading to increased cancer cell migration, invasion, proliferation and chemoresistance.[32] Cleavage of caspases is known to inhibit cell proliferation of cancer cells and induce apoptosis.[33] Hence, Poulaki et al.[34] investigated the potential of bortezomib as a treatment of retinoblastoma by inducing caspase-dependent apoptosis. It was found that bortezomib caused a stress response in the retinoblastoma cells, which led to inhibition of cancer cell proliferation and induced apoptosis at a clinically achievable concentration.[34] ARID1A is known to act as a tumor suppressor, the loss of which aids the proliferation, migration, and survival of cancer cells.[35] Liu et al.[36] investigated the expression levels of ARID1A by whole exome sequencing (WES) and found that it was found to be highly downregulated in retinoblastoma patients with high metastatic features and endophytic invasion.[36] MSH3 is yet another gene that was found to be highly downregulated in retinoblastoma and is associated with vitreous seeding.[37]

RNA sequencing analysis revealed a significant downregulation of DRAIC expression levels in retinoblastoma. DRAIC is known to play a crucial role in the proliferation and progression of cancer cells.[38] Restoring DRAIC RNA levels was found to significantly slow down the growth of retinoblastoma cells, indicating its importance in tumor evasion.[38]

DEK (qPCR and microarray analysis),[14] CRB1 (whole exome sequencing and Somatic copy number alterations [SCNA] profiling),[18] MIR181 (whole exome sequencing and SCNA profiling),[18] NUP205 (whole exome sequencing and SCNA profiling),[18] IL8 (microarray analysis),[18,39] IL6 (microarray analysis),[39] MYC (microarray analysis),[39] and SMAD3 (microarray analysis)[39] are some of the other genes that have been found to be overexpressed in retinoblastoma and usually correlate with retinoblastoma tumorigenesis towards an aggressive form.[26]

Similarly, several genes like CREBBP, HIST1H4H (MSK-IMPACT),[32] RELN (RNA sequencing),[32] RPTOR (MSK-IMPACT),[20] TERT (MSK-IMPACT),[20] MSH3 (MSK-IMPACT),[20] TSC2 (MSK-IMPACT),[20] ARID1A (single cell RNA sequencing),[20] CDK4 (qPCR),[40] BRAF (qPCR),[40] JAK1 (qPCR),[40] and ROCK1 (qPCR)[40] have been found to be downregulated in retinoblastoma tumors. Large deletions have been found in these genes, accounting for a potential oncosuppressor function.

An interactome of how these genes interact with each other is shown in Fig. 2.

Figure 2.

Interactome network of genes identified for its association with retinoblastoma (using STRING database)

Potential target gene identification from proteomic profiling of retinoblastoma

Only a limited number of studies have conducted proteomic profiling in retinoblastoma tumor or cell lines to identify potential proteomic targets and biomarkers. Using isobaric tags for relative and absolute quantitation (iTRAQ), 2-D difference gel electrophoresis (2D-DIGE) coupled with matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), and high-resolution mass spectrometry, several proteins have been identified, and alteration of these genes have been linked to clinical manifestations in retinoblastoma. Upregulation of IGF2BP1 protein levels were detected in retinoblastoma cells using liquid chromatography–mass spectrometry (LC–MS), and it is known to induce proliferation and migration of cancer cells.[41] The functional knockdown of IGF2BP1 revealed that reduced expression resulted in decreased proliferation and cell migration, leading to reduced tumor migration and induction of apoptosis in the retinoblastoma tumor cells. Therefore, IGF2BP1 could be a promising therapeutic target for retinoblastoma.[41] An important developmental transcription factor, SOX4, is known to regulate cell proliferation and stemness features under normal conditions, was found to be increased in retinoblastoma tumors by immunohistochemistry (IHC).[42] The increased expression was associated with tumors that had invaded the optic nerve as opposed to those without invasion. These findings suggest that SOX4 expression could be a valuable prognostic marker for retinoblastoma, particularly in cases involving optic nerve invasion.[42]

Another highly mutated gene at the proteomic level was B7H3, which was found to be highly expressed in retinoblastoma cells, as identified via IHC analysis, and is known to inhibit the functioning of T cells and to actively participate in tumor invasion.[43] The levels of B7H3 were found to be upregulated in poorly differentiated retinoblastoma compared to the moderate and well-differentiated tumors, indicating its association with massive choroid invasion. The study also suggests that B7H3 expression may downregulate the expression of T-cell responses, which could explain how B7H3 presence in retinoblastoma tumors promotes the development of targeted therapeutics and introduces immunotherapy approaches as a viable alternative for chemoresistant tumors.[43] PEDF levels have been found to be downregulated in retinoblastoma, as validated by mass spectrometry (MS).[44] When increased, PEDF can act as a tumor suppressor by delaying tumor growth, reducing VEGF levels, and inhibiting tumor cell invasion and migration.[44]

The nucleolin (NCL) protein has been known to be overexpressed in retinoblastoma, which helps in tumor proliferation, as identified via qPCR analysis. Subramanian et al.[45] used an FDA-approved NCL-targeting agent called NCL aptamer/AS1411 to reduce the proliferation of tumor cells and thus inhibit the proliferation and progression of retinoblastoma cells. The study showed that NCL aptamer led to cell cycle arrest, the downregulation of oncogenic microRNA (miRNA), and inhibition of cell proliferation in retinoblastoma cells.[45] Calcineurin overexpression has been known to induce hypoxic stress to cells promoting cancer cell proliferation, validated via flow cytometry. Therefore, cyclosporin A (CSA), a well-known calcineurin inhibitor, was tested against retinoblastoma cells in a study.[46] The study showed that CSA induced a dose-dependent antiproliferative and pro-apoptotic response in retinoblastoma cells by potently inhibiting NFAT-mediated reporter gene transcription of these tumor cells. Hence, these results demonstrate the functional integrity of NFAT/calcineurin signaling in retinoblastoma and the cytotoxic effects of CSA against it in retinoblastoma cells.[46]

Several small molecule inhibitors like histone deacetylase inhibitors (HDACi) are being extensively used for cancer therapy. To evaluate its potential utility in retinoblastoma treatment, the effects of MS-275 (an HDACi inhibitor), in combination with antineoplastic drugs like vincristine, etoposide, and carboplatin, were studied in vivo, which determined that it was a well-tolerated and highly effective treatment option for retinoblastoma.[47] MS-275 has shown promising results in phase I and phase II clinical trials against various lymphoid carcinomas and leukemias and can be considered for retinoblastoma as well.[48,49]

Similarly, several other proteins like chromogranin A (IHC),[41] Rac GTPase-activating protein 1 (IHC),[41] fetuin A (IHC),[41] midkine (IHC),[41] LRP1 High resolution mass spectrometry(HR-MS),[50] COMP (HR-MS),[50] TGB3 (HR-MS),[50] TLN (HR-MS),[50] FLNA (HR-MS),[50] OGN (HR-MS),[50] A1BG (IHC),[44,50] Serpin A1 (IHC),[44,44] ORM2 (IHC),[44] LRG1 (IHC),[44,44] CHI3L1 (IHC),[44,44] apolipoprotein A1 (western blotting),[51] transferrin (western blotting),[51] alpha-crystallin A (western blotting),[52] and CRABP2 (western blotting)[52] have been found to be upregulated in retinoblastoma.

These studies thereby provide a dynamic protein profile of retinoblastoma tumors that can provide clues to study the oncogenesis mechanism of retinoblastoma and identify possible potential biomarkers.

Potential gene identification from epigenetics profiling of retinoblastoma

Studies focusing on epigenetic aberrations have suggested that several epigenetic dysregulations hugely impact cancer pathways and contribute to the malignant progression of various tumors. The probability of epigenetic dysregulation in retinoblastoma has been explored through integrative epigenome analysis. Epigenetic mechanisms such as hypermethylation, gene silencing, and large deletions have been found to be involved in retinoblastoma progression. SYK protein levels have been found to be highly elevated in retinoblastoma tumors of orthotropic xenografts—identified via whole genome sequencing—compared to the human fetal retina as the protein does not play any functional role in developing the visual system.[53] Lentivirus Lenti-SYK-9 was used to decrease the SYK levels, which drastically increased apoptosis in retinoblastoma cells. Additionally, SYK inhibitors, such as BAY 61-3606 and R406, showed that SYK inhibition resulted in mitochondrial defects and morphologically visible cell death. When combined with topotecan, BAY 61-3606 significantly improved the outcome. These results suggest that SYK is a promising new target for treating retinoblastoma.[53] DNMTs such as DNMT1a and DNMT3a have been found to be highly overexpressed in poorly differentiated retinoblastoma tumors, as validated by MLPA analysis.[54] The study demonstrated that high expression levels of DNMT1a and DNMT3a are correlated with more aggressive disease and highly malignant phenotypes that are considered to be indicative of poor prognosis. Therefore, protein expression levels of DNMTs could be a reliable prognostic factor and offer the possibility of a new therapy for retinoblastoma treatment.[54,55] Similarly, presence of RASSF1A was found in a high proportion of retinoblastoma cases that were identified via qPCR analysis.[56] Loss of RASSF1A was observed as the most frequent epigenetic aberration in retinoblastoma, which was not observed in normal retinae or the non-malignant retinoblastoma cases, indicating its role in tumor metastasis.[56] MDM4—which is an inhibitor of p53 protein that disrupts the apoptotic pathway—has been found to be highly upregulated in retinoblastoma with the help of immunoblotting and immunoprecipitation.[24] MDM4 was targeted by a small molecule kinase inhibitor – CEP1347 and its effects were validated in retinoblastoma cells by Togashi et al.[57] CEP1437 inhibited the growth of MDM4 and reduced its expression while activating the p53 apoptotic pathway in retinoblastoma cell lines. This comes forth as a promising candidate for retinoblastoma therapy. Downregulation of HELLS protein (an epigenetic modifier) was also determined to be a key factor driving retinoblastoma tumorigenesis, with high expression levels being validated via qPCR as well as IHC (39). Although HELLS is involved in the cell differentiation and development of the retina, it is not essential for its maintenance once the major cell types have been formed. A HELLS knockdown study revealed a decrease in tumor burden and morbidity, hinting at its potential as a therapeutic target for retinoblastoma treatment.[58] Protein Kinase aberrations are known to be very common in cancers and PLK1 was found to be highly upregulated in retinoblastoma identified via RNA sequencing.[59] PLAK1 promotes the proliferation and progression of the retinoblastoma tumor and was targeted by ON-01910.Na, which demonstrated tumor-specific cytotoxicity, efficient drug permeability with multi-layer cell death, and cell-cycle arrest, all of which collectively resulted in the cell death of tumor cells.[59] Using a zebrafish model, Yang et al.[60] demonstrated how xanthatin selectively inhibited retinoblastoma cell proliferation by inducing cell cycle arrest and promoting apoptosis in tumor cells by directly targeting PLK1. HMGA2 is a known proliferative agent of cancer cells in retinoblastoma, which has been validated via qPCR.[60] Nalini et al.[61] developed a stable phosphorothioate-modified HMGA2 aptamer to block HMGA2 protein function and to induce cytotoxicity and inhibit retinoblastoma cell proliferation. Further in vivo studies might help understand the intraocular efficacy of HMGA2 aptamer in retinoblastoma cells.[61] Tigecycline is another FDA-approved antibiotic that has been tested in vitro and in vivo for its effects against retinoblastoma.[62] It showed significant growth inhibition and apoptosis in multiple retinoblastoma cell lines by causing excessive mitochondrial dysfunction and oxidative damage to the retinoblastoma tumor cell. The drug effectively inhibited angiogenesis, cell migration, invasion, proliferation, and survival of retinoblastoma tumors.[62]

Several genes were identified using the multiplex ligation-dependent probe amplification (MLPA) and methylation-specific (MS)-MLPA analysis like T3A,[54] MSH6, CD44, PAX5, GATA5, TP53, VHL, and GSTP1.[63] Additionally, deletions of the oncosuppressor genes, namely, TP53, CDH13, GATA5, CHFR, TP73, and IGSF4, were also identified in the tumors using MLPA analysis.[63] Similarly, epigenetic silencing of MGMT, CDK1, BUB1, CCNB2, CCNB1, TOP2A, RRM2, KIF11, KIF20A, NDC80, and TTK, were identified via bioinformatic analysis of retinoblastoma tumors that were associated significantly with poor survival outcomes based on patient survival analysis.[64] With the help of LC–MS, H2AFX and SIRT1 were identified as two hyperphosphorylated stress response proteins in retinoblastoma along with hyperphosphorylated protein kinases like BRD4, WNK1, and CDK1.[65]

To further the understanding of epigenetic mechanisms that regulate retinoblastoma progression, Corson and Gallie[15] compiled available literature and identified several aberrantly methylated genes, including CASP8, and MLH1, associated with retinoblastoma progression. The genetic, transcriptomic, and proteomic alterations in several genes have been explained in Fig. 3.

Figure 3.

The figure represents the various genetic, transcriptomic, and proteomic aberrations observed in several driver genes that are known to promote tumor progression in retinoblastoma. Red (down) – chromosomal loss, deletions, methylation, and downregulation. Green (up) – chromosomal gain, insertion, amplification, and upregulation, White (ND) – Not Determined. (RB1 chromosomal loss, RB1 deletion, RB1 aberrant methylation, RB1 transcriptomic regulation, RB1 proteomic regulation.[66] MYCN chromosomal gain,[67] MYCN transcription regulation, MYCN proteomic regulation,[68] KIF14 chromosomal gain, KIF14 transcriptomic and proteomic regulation,[23] E2F3 chromosomal gain, E2F3 transcriptomic regulation,[69] DEK chromosomal gain and transcriptomic regulation,[14] BCOR mutation,[20] CREBBP chromosomal gain,[3] CREBBP INDELs, MDM2/MDM4 chromosomal gain, MDM2/MDM4 transcriptomic regulation,[70] TP53 deletion mutation and aberrant methylation,[63] SOX4 chromosomal gain,[18] CDH11/CDH13 chromosomal loss, mutation, and transcriptional regulation,[71] CRABP1 proteomic regulation,[51] MGMT aberrant methylation, MGMT transcriptomic regulation,[72] OTX2 transcriptomic regulation[27]

A list of all the genes identified via different multi-omics technologies have been summarized in Table 2. Different molecular aberrations observed at various stages of retinoblastoma oncogenesis have been explained with the help of a schematic diagram in Fig. 4. Insights from WES analysis of the tumor in two unilateral retinoblastoma cases and correlation with the clinical course of the disease at our hospital are provided in Figs. 5 and 6.

Table 2.

Genes involved in retinoblastoma tumorigenesis identified via multi-omics technologies

Sample type and size	Methodology used		Gene identified	Geographical region of the study and reference
	Primary screening	Validation
26 retinoblastoma tumor tissue samples	CGH	Southern blotting	LRRN5	Marburg, Germany Human Genetics, 2001[22]
68 retinoblastoma tumor tissues (40 bilateral and 28 unilateral) and WERI-RB1, Y79 cell lines	Hypermethylated genes previously identified: RASSF1A gene[73] and MGMT gene[72]	Epigenetic analysis: MSP and bisulfite sequencing	RASSF1A and MGMT	Hong Kong, China and New York, USA Invest Ophthalmol Vis Sci, 2002[56]
76 retinoblastoma tumor tissues and Y-79 and WERI-RB1 cell lines	qRT-PCR and microarray analysis	qRT-PCR and western blotting	E2F3 and DEK	Essen, Germany Oncogene, 2005[14]
Weri-Rb-1, and NCC-RbC-51 cell lines	Microarray and bioinformatic analyses	qRT-PCR	IL8, IL6, SMAD3, and MYC	Chennai, India Experimental Eye Research, 2020[39]
8 Retinoblastoma tumor tissues	Proteomic profiling and immunoblotting	IHC	B7H3	Chennai, India Scientific Reports, 2020[43]
30 retinoblastoma tumor tissue samples	RT-PCR	IHC	KIF14	Chennai, India Investigative Ophthalmology and Visual Sciences, 2007[23]
128 tumor tissue samples from enucleated eyes of RB patients (20/128 with retinoma)	Gene-specific FISH and qPCR	Immunostaining and RT-PCR	NGFR, KIF14, MYCN, CDH11, DEK, and E2F3	Toronto, Canada and Philadelphia, USA Human Molecular Genetics, 2008[74]
29 retinoblastoma tumors (well, moderately, and poorly differentiated)	2-DE and MALDI-TOF MS	IHC	Apolipoprotein A1, transferrin, CRABP2, α-crystallin A, recoverin, and peroxiredoxin	Chennai, India Proteomics Clinical Applications, 2010[52]
4 primary retinoblastoma tumors	WGS	Orthotopic xenografts from 4 primary human retinoblastoma tumors and 3 RB cell lines (Y79, WERI-RB1, RB355)	SYK	Tennessee, USA Springer Nature, 2012[53]
12 retinoblastoma tumor samples (5 bilateral and 7 unilateral)	MS-MLPA	Statistical analysis	MSH6, CD44, PAX5, GATA5, TP53, VHL, and GSTP1, RB1, MGMT, CDKN2, CDH13, CHFR, TP73, and IGSF4.	Siena, Italy Pathology and Oncology Research, 2012[63]
21 retinoblastoma tumor tissue samples	Microarray analysis using CGH	qRT-PCR	ASPM, CENPF, KIF14, NUF2, NEK2, CKS1B, NR5A2, KIFC1, NUP153, E2F3, UHRF1BP1, FOXP4, SOX4, TEAD3, DAXX, DEK, EHMT2, JARID2, RUNX2, PRPF4B, RBL2, USP10, FBXO31, NAE1, CHD9, PRMT7, BBS2 and RPGRIP1L	Birmingham, UK British Journal of Cancer, 2013[75]
94 retinoblastoma tumors obtained from enucleated eyes of malignant RB patients	SNP 6.0 analysis	qRT-PCR	RB1, BCOR, MYCN, OTX2	Memphis, USA Oncotarget, 2014[17]
45 primary retinoblastoma tumor samples	WES and SCNA profiling, and 265 tumor samples obtained from 11 studies	High-resolution microarrays on 8 retinoblastoma cell lines (RB1021, RB383, RB247, RB191, RB176, WERI-RB1, and Y79)	CRB1, NEK7, MIR181, SOX4, DEK, and NUP205	Amsterdam, The Netherlands PLos One, 2016[18]
5 primary retinoblastoma tissues	Proteomic analysis: iTRAQ and MS	IHC	Chromogranin A, Rac GTPase-activating protein 1, Fetuin A, Midkine, and IGF2BP1	Chennai, India Clinical Proteomics, 2016[41]
Aqueous humor (AH) samples collected from 10 patients with group D retinoblastoma	Proteomic analysis: iTRAQ technology	IHC	A1BG, Serpin A1, ORM2, LRG1, CHI3L1, PEDF, and STRA6	Beijing, China Oncology Letters, 2017[44]
30 RB tumor tissue specimens from enucleated eyes	Proteomic analysis: 2D-DIGE coupled with MALDI-TOF MS	qRT-PCR and western blotting	APOA1, RBP3, GFAP, CRYAA, TF, CRABP1, and SAG	Chandigarh, India Journal of Proteomics, 2017[51]
Retinoblastoma cell lines – RB176, RB177, RB212, RB214, RB216, RB217, RB218, Y79, and WERI-RB1 Orthotopic xenografts (8-week-old male athymic mice)	Western blot analysis	qRT-PCR and knockdown study analysis	MDM2	Los Angeles, California USA Oncogene, 2017[24]
5 Retinoblastoma tissues	Proteomic analysis: iTRAQ technology and orbitrap mass spectrometry	IHC	LMNB1 and TFRC	Karnataka, India Proteomics Clinical Applications, 2018[76]
8 primary retinoblastoma tumor samples (undifferentiated and differentiated)	Microarray analysis using (HTA2.0)	qRT-PCR	CDK4, BRAF, JAK1, and ROCK1	Mexico City, Mexico Journal of Cancer Research and Clinical Oncology, 2020[40]
Y79/EDR cell lines	RNA sequencing	qRT-PCR	STC1, HK2, ARHGAP9, RELN, DDIT4, PFKB4, and HIST1H4H	Beijing, China BMC Ophthalmology, 2020[32]
Cell lines (HSJD-RBT1, RBT2, RBT5, RBT14 and HSJD-RBVS1, RBVS3, RBVS10)	Proteomic analysis: HR-MS	Gene Ontology analysis	LRP1, COMP, TGB3, TLN, FLNA, and OGN	Barcelona, Spain and Rome, Italy Cancers, 2020[50]
6 NCBI GEO datasets (206 RB samples total)	Data processing and analysis of multi-omics microarray data by integrative bioinformatic analysis	DNA methylation and microRNA analysis	CDK1, BUB1, CCNB2, CCNB1, TOP2A, RRM2, KIF11, KIF20A, NDC80, and TTK	Hubei, China Medicine, 2020[64]
Retinoblastoma tumor tissues	Phosphoproteomic profiling	MS	H2AFX, SIRT1, BRD4, WNK1, and CDK1	Chennai, India Molecules, 2018[65]
2 retinoblastoma samples from enucleated eyes	scRNA-seq and multiresolution network-based analysis	UBE2C-knockdown study on orthotopic retinoblastoma model in BALB/c nude mice	UBE2C	Shanghai, China Cell Death and Disease, 2021[26]
83 primary retinoblastoma tumor samples (after primary/secondary enucleation)	MSK-IMPACT clinical NGS	FACETS	BCOR, RPTOR, TERT, MSH3, ARIDIA, MYCN, TSC2, and CREBBP.	New York, USA Cancers, 2021[20]

CGH=comparative genomic hybridization; qRT-PCR=real-time quantitative reverse transcription polymerase chain reaction; MSP=methylation-specific polymerase chain reaction; FISH=fluorescence in-situ hybridization; PFPE=PAXgene-fixed paraffin embedded; 2-DE=two-dimensional gel electrophoresis; qPCR=quantitative polymerase chain reaction; MALDI-TOF MS=matrix-assisted laser desorption/ionization-time of flight mass spectrometry; WGS=whole genomic sequencing; MS-MLPA=methylation-specific multiplex ligation probe assay; SNP analysis=single nucleotide polymorphism analysis; WES=whole exome sequencing; SCNA=somatic copy number alterations; iTRAQ=isobaric tags for relative and absolute quantitation; MS=Mass spectrometry; HR-MS=High-resolution mass spectrometry; scRNA-seq – single-cell RNA sequencing; MSK-IMPACT=Integrated Mutation Profiling of Actionable Cancer Targets; FACETS=fraction and allele-specific copy number estimates from tumor sequencing; NGS=next generation sequencing

Figure 4.

Molecular alterations at different stages of retinoblastoma tumorigenesis. (A) Normal Retina – developing with normally regulated pathways and gene expression along with a proper functioning DNA repair mechanism.[77] (B) Retinoma – development of a cluster of benign cancer cells with differential expression of several genes[16] (C) Retinoblastoma – cancer cells begin to grow uncontrollably with an overexpression of several oncogenes and aberrant genetic alterations in others[20] (D) Malignant progression – when the cancer begins to metastasize to other parts of the body from the primary tumor site with differential gene expression[32,3] (E) Vitreous seeding of the tumor where small parts of tumor reach the vitreous humor and is characterized by differential expression of several proteins[50]

Figure 5.

Left eye aggressive unilateral extraocular retinoblastoma in a 9-year-old child. (a) She was found to have RB1 wild type; however RB1 promoter was 100% methylated, which caused the downregulation of RB1. MRI orbits show extraocular extension with diffuse optic nerve thickening reaching till chiasma. (b) She underwent enucleation after chemoreduction. Apart from RB1, several other mutations were found in the tumor via WES of tumor DNA such as ARHGAP9 mutation (p.Cys130Arg, c.388T>C, Tgc/Cgc) and MSH3 mutation (p.Ser420Gly, c. 1258A>G, Agc/Ggc), the downregulation of both of which is known to be involved in metastasis, optic nerve invasion, and chemoresistance. Similarly, we found mutation in the E2F3 gene (p.Ser76Gly, c.226A>G, Agc/Ggc) which is found to be highly overexpressed in retinoblastoma and helps in the growth and proliferation of tumor cells

Figure 6.

Unilateral left eye aggressive retinoblastoma in a 4-year-old child (a) with proximal optic nerve enhancement and thickening at presentation (b). Left phthisis bulbi following chemoreduction (c) and MRI showing resolution of optic nerve thickening (d). Patient underwent enucleation in left eye and developed massive orbital recurrence (e and f) 2 months after enucleation, indicating an aggressive tumor. The tumor tissue was isolated from the enucleated eyeball, from which the DNA was isolated to perform WES. We identified a stop gain mutation (p.Arg255*, c.763C>T, Cga/Tga) in the RB1 gene which validates the loss of RB1 in aggressive retinoblastoma tumors. In addition to this, ARID1A mutation (p.Arg1383Gln, c.4148G>A, cGg/cAg) was also identified via WES whose loss is associated with highly metastatic features in retinoblastoma

Pathways Associated with RB1 Inactivation in Retinoblastoma

Apart from the RB1 loss,[3] various genetic aberrations are involved in retinoblastoma progression that assist the dysregulation of various signaling pathways. The retinoblastoma protein (pRb)—which acts as a central regulator of cell cycle entry and inhibits transcription by binding to E2F transcription factor—aids in limiting the cell cycle progression.[78] Cell cycle progression is inhibited by the loss of RB1 gene and concurrent inhibition of SKP2, leading to the hinderance of stabilization of p27 expression.

In retinoblastoma, amplification of MDM2 and MDM4 genes enhances tumor progression by suppressing the p53-mediated apoptotic pathway. This leads to a detectable loss of p14^ARF proteins, thereby promoting tumor proliferation and survival. Following the loss of RB1 gene, SKP2 acts as a lethal gene. This causes p27 to be downregulated and SKP2 to be overexpressed, thereby dysregulating the entire TRβ1/2 signaling pathway.[79] These events collectively enhance retinoblastoma cell growth and development.[80] Consequently, dysregulation of pRb disrupts a number of pathways involved in cell cycle progression and apoptosis of tumor cells in retinoblastoma. Various deregulated pathways that contribute to tumor progression in retinoblastoma have been explained in Fig. 7. A brief summary of some small molecule inhibitors used against retinoblastoma is provided in Table 3.

Figure 7.

Schematic diagram of various dysregulated pathways and differentially expressed genes that promote tumor proliferation and development in retinoblastoma identified via multi-omics analysis of tumor samples and retinoblastoma cell lines[20,78]

Table 3.

Small molecule inhibitors for retinoblastoma

Name of the drug	Target gene	Cell line tested	In vivo model tested	Reference
MLN4924 (Pevenedistat)	NAE/SKP2	RB1021, RB3823, RB381, RB522, RB247, RB3535S, WERI-RB1 and Y79	3–4-week-old NOD-Scid mice	[81]
Nutlin-3A	p53-HDM2	Y79, WERI-RB-1, and MDA-MB-435	ND	[82]
CEP1347	MDM4	Y79, WERI-RB-1, NCC-RbC-54, NCC-RbC-60, and NCC-RbC-83	ND	[57]
ON-01910.Na	PLK1	WERI-RB1, Y79, and ARPE-19	4–6-week-old female, BALB/c (nu/nu) immunodeficient mice	[83]
HMGA2 (high-mobility group A2) aptamer	HMGA2	Y79, WWERI-RB1 and MIO-M1	ND	[61]
Bortezomib	NF-Kb	Y79 and WERI-RB1.	ND	[34]
Tigecycline	Wnt/β-catenin	Y79, WERI-RB-1, and RB116	NOD/SCID mice	[62]
Lenti-SYK-9	SYK	Y79, WERI-RB1, and RB355	immunocompromised SCID mice	[53]
Nucleolin-aptamer (NCL-APT)	NCL	Y79 and WERI-RB1	5-week-old female nude mice	[84]
pri-miRNA-17∼92 aptamer (pri-apt)	miR-17, miR-18a, and miR-19b	WERI-RB1 and Y79	ND	[85]
CD44-thioaptamer, A15 aptamer, and A12 and A35 aptamers	Thioaptamer targeting CD44, A15 aptamer targeting CD133, and A12 and A35 aptamers targeting ABCG2 CSC marker	Y79 and WERI-RB1	ND	[86]
EpDT3 – RNA Aptamer conjugated with Doxorubicin	EpCAM	Y79 and WERI-RB1	ND	[87]
RNA G-quadruplex aptamer	pRB	H460, RB-knockout H460, HEK293, MCF7, and U87MG	ND	[88]
HDAC inhibitor	HDAC	ND	Transgenic mouse and rat ocular xenograft models	[47]
Cyclosporin A	NFAT signaling/Calceneurin	Y79 and WERI-RB1 cell lines	ND	[46]
Xanthatin	PLK1-mediated G2/M pathway	Y79 and WERI-RB1 cell lines	Zebrafish model	[60]

NAE=NEDD8 activating enzyme; SKP2=S-phase kinase associated protein-2; SYK=spleen-associated tyrosine kinase; NCL=nucleolin; CSC marker=cancer stem cell marker; EpCAM=epithelial cell adhesion molecule; HDAC=histone deacetylase; ND=not defined

Futuristic therapies and models for retinoblastoma

Based on the concepts of immunotherapy, oncolytic viruses, and hydrogel technology for delivery of targeted drugs, various innovative therapies have been introduced against retinoblastoma. Additionally, to further validate the outcomes of these targeted retinoblastoma therapeutics and test its clinical efficacy, several molecules are undergoing clinical trials. However, since the disease is rare and complex, the number of controlled clinical trials and pre-clinical studies addressing retinoblastoma treatment have been limited.

• Oncolytic adenovirus targeting retinoblastoma – A clinical study was conducted by Pascual-Pasto et al.[89] to evaluate the oncolytic activity of oncolytic adenovirus (VCN-01) and to test the safety and activity of it as treatment for chemo-refractory retinoblastoma in pediatric patients. VCN-01 is an oncolytic adenovirus that has been developed to replicate in tumor cells which have an abundance of free E2F-1, Increased level of free E2F-1 is a consequence of a distorted RB1 gene molecular pathway. With the help of this study, it was proven that VCN-01 was active, safe, tumor selective, and clinically translatable against retinoblastoma.

• CAR-T cell therapy for retinoblastoma – Adaptive T -cell immunotherapy has received a lot of attention as a possible therapy for many cancers, including retinoblastoma. Andersch et al.[90] conducted a preclinical study to find the efficacy of CAR-T cell therapy targeting CD171 and GD2 antigens that were specifically associated with retinoblastoma tumors. Their study showed that CD171- and GD2-specific CAR-T cells displayed potent activation of apoptosis, thereby generating a strong anti-tumor response against retinoblastoma cells. Furthermore, sequential switching of the antigen specificity in CAR-T cell therapy provides beneficial results for the killing ability of tumor cells. To overcome the limitations of systemic administration of CAR-T cells in retinoblastoma, Wang et al.[91] generated GD2-specific chimeric antigen receptor T lymphocytes (GD2.CAR-Ts) against GD-2 and studied their expression in vivo. In this preclinical model, they combined effector CAR-T cells with T-cell growth factor IL-15 and encapsulated them in a chitosan–polyethylene glycol (PEG) injectable hydrogel. The results indicated that GD2.CAR-Ts showed potent cytotoxic activity against retinoblastoma cell lines in vitro, where IL-5 improved the T-cell lifespan and the injectable hydrogel enhanced its biodistribution.

• iPSC-derived 3-D retinal organoid models for retinoblastoma – To study the model in vivo, Norrie et al.[92] developed iPSC-derived 3-D retinal organoids from retinoblastoma patients and injected them into the eyes of immunocompromised mice. The retinoblastoma tumor formed from these organoids in mice were indistinguishable to human retinoblastomas in terms of their cellular, molecular, and genomic features. Therefore, this organoid model based on patient-derived iPSCs contributes to our understanding of the cellular and molecular origins of retinoblastoma tumorigenesis mechanism post Rb1 inactivation.

• Hematopoietic stem cell therapy for retinoblastoma – In an interesting case study, Rastogi et al.[93] reported a case of treating a relapsed metastatic retinoblastoma child after hematopoietic stem cell transplantation (HSCT) and chemotherapy treatment. The therapy was successful in treating abdominal metastatic lesions but did not prevent CNS relapse metastasis. The findings suggest that HSCT therapy could be performed on retinoblastoma cases of primary metastasis. However, additional therapy may be needed to address the CNS involvement. To overcome this issue, various studies have reported the use of drugs like topotecan or thiotepa.[94] Stem cell therapy is being used extensively in cancer treatment and has great potential to be incorporated in retinoblastoma treatment on further research.

• Anti-HDM2 peptide delivery to retinoblastoma cells using gold nanoparticles – Biocompatible gold nanoparticles (GNPs) were used to deliver anti-HDM2 peptides to retinoblastoma cells.[95] In vitro analysis showed that these peptides function by arresting the retinoblastoma cells at the GM2 phase of the cell cycle, thereby leading to p53-induced apoptosis.

Conclusion

Recent developments in our understanding of the molecular landscape of retinoblastoma has equipped us with a valuable list of potential targets for targeted therapy, thanks to the success of multi-omics technologies like NGS and proteomics. These targets have high therapeutic potential in targeted therapy against retinoblastoma, and various small molecule inhibitors are being investigated for it. Effective clinical translation of these agents remains an ongoing challenge. With inadequate outcomes being observed in the clinical trials of targeted therapeutics, further studies are required to close the gap between the preclinical and clinical efficacy of these therapies.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.