To minimize risk of tumor growth outside of the eye, retinoblastomas have historically not been biopsied and characterized as they grow. Ruling out such tumor biopsy, patient care decisions are generally made without understanding the tumor cell characteristics or the immediate and long-term risks. Existing methods to classify retinoblastomas based on clinical features can help guide treatment. Criteria including tumor location within the eye, cell invasion and metastasis, and tumor seeding can help predict the likelihood of saving a child’s eye but there is no perfect algorithm to determine when to remove an eye and when to attempt therapy. Laboratory research on molecular features of retinoblastoma can help answer questions inaccessible to clinical examination.
In the current era there is a need to harness large datasets to better understand the biology of childhood cancers. The Childhood Cancer Data Initiative (CCDI) was proposed in 2019 to identify genomic alterations in pediatric cancer and to match patients with clinical trials. However, many key goals of the CCDI do not apply to retinoblastoma because the tumors can’t be biopsied. Additionally, many basic questions of retinoblastoma research can’t be answered with current approaches. Why do some eyes respond to treatment while others persist? What is the molecular basis for tumor relapse? What is the likelihood of eye salvage or long-term visual function for an individual patient? These questions can’t be directly answered without molecular biomarkers that are inaccessible because the cancer can’t be biopsied.
Several years ago, to address this need, our group developed a liquid biopsy of retinoblastoma that could be collected during treatment. Berry et al. demonstrated that aqueous humor (AH) is a source of cell-free tumor DNA (cfDNA). Aqueous humor provides direct access to cast off tumor DNA and is an ideal liquid biopsy for retinoblastoma. The team’s first study was published in JAMA Ophthalmology1 and soon proposed the clinical utility of aqueous humor as a liquid biopsy in 2020 to identify pathogenic hits in the RB1 gene2, detect prognostic genomic biomarker (chr6p gain and MycN amplification) that predict eye preservation3, and monitor treatment response with tumor fraction estimation4.
This work developed as a collaboration between researchers at the Vision Center at CHLA and the Convergent Sciences Institute at USC with institutional support from The Saban Research Institute. The group has also benefited from expertise at the Center for Personalized Medicine at CHLA. Now, with the epigenetics expertise of the laboratory of Gangning Liang at USC, the AH liquid biopsy platform has expanded to enable epigenetic analysis of proliferating retinoblastoma tumors.
Epigenetic alterations such as DNA methylation that impact gene expression apart from DNA mutation can influence tumor growth and evolution. For example, in many retinoblastomas, methylation of the RB1 promoter is the first step in formation of tumors. Additionally, epigenetic deregulation of cancersignaling pathways may impact disease progression. There have been multiple studies on DNA methylation in retinoblastoma, but these were done on tumor tissue only. Until now, the clinical impact of epigenetic changes in ongoing retinoblastoma cases was unknown. This report from Li and Xu et al.5 is the first to characterize tumor methylation profiles of RB tumors in vivo using the AH liquid biopsy. The authors derived tumor methylation signatures from sparse,
fragmented aqueous humor cell free DNA (cfDNA) obtained at diagnosis and throughout therapy to establish tumor methylation of ongoing retinoblastoma cases. They further found that AH methylation profiles closely mirror methylation patterns in matched RB tumors with distinct methylation patterns (clusters A and B) corresponding to tumor aggressiveness (subtype 1 and 2) and identified a set of genes whose methylation may be highly predictive of future disease. Such discoveries may allow the identification of new targets for retinoblastoma treatment. However, further in vitro and in vivo studies are needed to confirm the mechanisms of epigenetic regulation and roles of the novel genes identified in this study.
Retinoblastoma research in vitro has revealed the basis of tumor formation and helped define patterns of tumor progression. With the development of a liquid biopsy of retinoblastoma, there is now a potential to put decades of research on RB tumors into a clinically impactful assay for patients in the absence of tumor tissue. The study by Li and Xu et al. examines stable alterations in DNA methylation by high-throughput array technologies. By expanding the capacity of AH analyses to capture methylation status, the role of epigenetic alterations during disease can be defined clinically. The identification of a differentially methylated gene signature present in retinoblastoma and in matched AH cfDNA opens new horizons for precision medicine in retinoblastoma. These exciting findings call for validation through expanded studies to impact the lives of young patients with retinoblastoma.
- Berry, J. L. et al. Potential of aqueous humor as a surrogate tumor biopsy for retinoblastoma. JAMA Ophthalmol. 135, 1221–1230 (2017).
- Polski, A. et al. Variability in retinoblastoma genome stability is driven by age and not heritability. Genes. Chromosomes Cancer 59, 584–590 (2020).
- Berry, J. L. et al. Genomic cfDNA analysis of aqueous humor in retinoblastoma predicts eye salvage: the surrogate tumor biopsy for retinoblastoma. Mol. Cancer Res. 16, 1701–1712 (2018).
- Wong, E. Y. et al. Inter-eye genomic heterogeneity in bilateral retinoblastoma via aqueous humor liquid biopsy. Npj Precis. Oncol. 5, 1–6 (2021).
- Li, H.-T. et al. Characterizing DNA methylation signatures of retinoblastoma using aqueous humor liquid biopsy. Nat. Commun. 13, 5523 (2022).