The study by Gregersen et al1 on patients with retinoblastoma used data obtained from the Danish Ocular Oncology Group database and the Danish Cancer Registry, collecting long-term data on all 323 patients with retinoblastoma diagnosed in Denmark from 1943 to 2013. The scope of this research is ambitious.
The main issues this study sought to address were the incidence of second primary cancers in the Denmark retinoblastoma population and whether there was a difference in incidence between patients with heritable vs nonheritable retinoblastoma. The data were also stratified by treatment modality to examine the role that treatment played in the incidence of second primary cancer among these patients.
This study also offers detailed background information on the genetic differences between heritable and nonheritable forms of retinoblastoma as well as the evolution of retinoblastoma treatment. In this commentary, I touch upon the pathophysiologic characteristics and clinical presentation of the disease.
The median age at retinoblastoma diagnosis is 18 to 20 months. Children with unilateral disease are diagnosed at an average age of 24 months, while those with bilateral disease present at an average of 12 months. A total of 95% of all patients with retinoblastoma are diagnosed before age 5 years. Retinoblastoma typically presents as leukocoria. Thus, any observed leukocoria in a pediatric population should prompt a referral and evaluation by an ocular oncologist.2
Retinoblastoma is caused by a variation in both alleles of the RB1 gene, which is a tumor suppressor gene that restricts the cells’ progression from G1 to S phase. The 2-hit model has been proposed to distinguish heritable and nonheritable forms of retinoblastoma. The first hit occurs due to a germline variation in the first RB1 gene. The second hit occurs later in development and is somatic.2 In a review of 1068 cases of unilateral nonheritable retinoblastoma, 2.7% of the cases were associated with no variation. One half of the cases with no RB1 variation were associated with MYCN proto-oncogene amplification.3
Retinoblastoma can become very large, filling and destroying the entire eye. Metastatic disease typically occurs within the first 12 months of presentation with retinoblastoma. The most common route of metastatic spread is through the optic nerve into the central nervous system. However, the tumor can also spread via the choroid into the sclera and into the orbit. Subarachnoid and cerebrospinal fluid spread is also possible.2
The prognosis for eye salvage becomes worse with progression of the disease. Mortality for metastatic disease outside the orbit is greater than 50%. However the most common cause of death in patients with retinoblastoma is a second primary cancer,2 which Gregersen et al1 address.
Gregersen et al’s1 study provides a useful contribution to the work on pediatric retinoblastoma. However, the major point of contention is the study’s global relevance because of the limited target population. Since this study focuses solely on a Danish population, a question arises on how generalizable the results are to the rest of the world given different geographic demographic makeups. Publicly available data suggest that pediatric cancer rates between the US and Denmark are similar: 2016 data from the National Cancer Institute Surveillance, Epidemiology, and End Results program reports the incidence of pediatric cancer to be 19.3 per 100 000 US children,4 and 2014 data from the Danish Childhood Cancer Registry reports the incidence of childhood cancer to be between 15.5 and 22.2 per 100 000 Danish children.5
However, the US has a much larger African American and Hispanic population than Denmark. In a recent study, among a cohort of 67 061 pediatric cancer cases, 52.5% of the children were non-Hispanic White, 28.7% were Hispanic of any race, 10.6% were non-Hispanic Black, 7.4% were non-Hispanic Asian or Pacific Islander, and 0.9% were non-Hispanic American Indian/Alaskan Native.6 Another study on cancer in children reported that, relative to White children, Black children had a 28% decreased risk of cancer, while both Asian and Hispanic children had an approximate 15% decreased risk. Children of mixed White/Black ancestry also were at decreased risk. However, estimates for mixed White/Asian and White/Hispanic children did not differ from those of White children.7
Given these findings, there is a question on how generalizable the findings of Gregersen et al1 are to the pediatric population in the US and other countries. I welcome the research community to address this possible confounding variable in future study.
Published: October 22, 2020. doi:10.1001/jamanetworkopen.2020.22569
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Ratna H. JAMA Network Open.
Corresponding Author: Haran Ratna, MD, MPH, Internal Medicine Department, Elmhurst Hospital Center, 79-01 Broadway, Elmhurst, NY 11373 (firstname.lastname@example.org).
Conflict of Interest Disclosures: None reported.