Figure. Flowchart demonstrating outcome of genetic evaluation for retinoblastoma (RB) patients evaluated in our program from 2001 to 2008.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Dhar SU, Chintagumpala M, Noll C, Chévez-Barrios P, Paysse EA, Plon SE. Outcomes of Integrating Genetics in Management of Patients With Retinoblastoma. Arch Ophthalmol. 2011;129(11):1428–1434. doi:10.1001/archophthalmol.2011.292
Author Affiliations: Departments of Molecular and Human Genetics (Dr Dhar), Pediatrics (Drs Chintagumpala and Plon and Ms Noll), and Ophthalmology (Dr Paysse), and Texas Children's Cancer Center (Drs Chintagumpala and Plon and Ms Noll), Baylor College of Medicine; Retinoblastoma Center of Houston (Drs Chintagumpala, Chévez-Barrios, and Plon); and Department of Pathology, The Methodist Hospital (Dr Chévez-Barrios), Houston, Texas.
Objective To present the outcome of a comprehensive team approach to provide genetic evaluation and testing for a large cohort of children diagnosed with retinoblastoma.
Methods The multidisciplinary team included pediatric oncologists, an ophthalmologist, an ophthalmic pathologist, a geneticist, and genetic counselors. Retrospective data from 8 years included 90 initial evaluations, of which 81 probands were diagnosed with retinoblastoma (34 bilateral and 47 unilateral) and 9 were evaluated because of a positive family history.
Results Genetic testing was accomplished equivalently in bilateral and unilateral cases in 51 of 81 patients (63%). In 5 of 30 patients (17%), with unilateral disease an RB1 mutation was identified in peripheral blood samples. In another 7 of 30 patients (23%), mutation analysis confirmed the occurrence of sporadic retinoblastoma. Overall, genetic testing of 48 at-risk family members from 21 families revealed 6 individuals positive and 42 negative for the familial mutation.
Conclusions Our study emphasizes that genetics can be incorporated into the management plan of all retinoblastoma patients using a team approach to ensure timely evaluations and appropriate counseling. Genetic evaluations improved risk prediction for patients and family members as well as prevented overutilization of clinical screening tests, which had potential morbidity for relatives documented to not carry an RB1 mutation.
Quiz Ref IDRetinoblastoma (RB) is a rare malignant tumor of the developing retina, with an incidence of 1 case in 16 000 to 18 000 live births.1-4 It is the most common primary ocular malignant neoplasm seen in children.5 Bilateral RB, found in 40% of cases, is caused by an inherited or de novo germline mutation in the RB1 gene (OMIM 180200).6 Sixty percent of patients presenting with unilateral RB result from either somatic mutation during development or, in approximately 15% of these patients, from hereditary mutations.7 Ten percent to 15% of all RB patients (unilateral and bilateral combined) report a positive family history of RB.8
Molecular genetic testing of the RB1 gene has been clinically available for more than 15 years, and sensitivity has continued to increase as techniques to analyze the RB1 gene have improved. Laboratories report the ability to detect a germline mutation in 90% to 95% of individuals with bilateral RB, depending on the methods used.9
For individuals with bilateral disease or a known family history of RB, genetic testing is performed directly on blood samples to identify heterozygous RB1 mutations. In unilateral cases of RB (with no family history), the testing strategy begins with analysis on fresh-frozen tumor DNA when available after enucleation. Tumor DNA may show either 2 RB1 mutations or 1 mutation on 1 allele and loss of heterozygosity or inactivation through methylation on the other allele. Peripheral blood samples can then be tested for any RB1 mutation found in the tumor to determine whether this represents a hereditary (germinal mutation) case.10
The primary rationale for genetic testing in RB is to aid in determining the risk of subsequent cancers (both subsequent RB and other primary neoplasia) in the affected child and the risk of RB in other family members, eg, siblings and offspring. Surveillance protocols for contralateral RB have been clearly established for children with unilateral disease.8 At-risk family members found to carry the same RB1 mutation as the proband also benefit from early and intensive screening for RB. At our institution, children with a known RB1 mutation have an eye examination every 3 to 6 weeks until age 1 year, and then every 3 months until age 3 years, and then every 6 months until age 6 years. Young children require examinations under anesthesia (EUA). Conversely, those who test negative for a familial RB1 mutation do not require ophthalmologic screening, with a significantly reduced health care–related cost and decreased potential for screening-related morbidity.9-11 Similarly, the identification of a germline RB1 mutation in the proband improves genetic counseling for the parents with regard to their risk of having subsequent children with RB as well as provides the opportunity for preimplantation or prenatal genetic testing.12,13 Conversely, for those children with unilateral RB in whom testing of tumor and blood samples demonstrate 2 somatic mutations in the tumor and no mutation in blood, ie, a documented sporadic case, the substantially reduced risk of RB in other family members and the reduced risk of a second malignant neoplasm for the proband is very important information.14
Although recommendations for genetic evaluation of RB patients have been published, few reports document the success of implementing the recommendations.15,16 In this article, we describe the results of a comprehensive approach to provide genetic testing and counseling to all RB patients, using a multidisciplinary team of pediatric oncologists, ophthalmologists, a geneticist, and genetic counselors at a large children's hospital. We present data from an 8-year period demonstrating the importance of genetics for the overall management of the patient with RB.
In this retrospective chart review, we included 81 children diagnosed with RB at Texas Children's Hospital (Houston) and 9 children with a family history of RB in first-degree relatives who were the first family member evaluated. Evaluation of subsequent family members of the RB patients is described in the “Clinical Pathway” subsection. Patients were identified from the patient log of the Baylor Cancer Genetics Clinic and the clinical database of the Retinoblastoma Service of the Texas Children's Cancer Center. They were initially seen for genetic evaluation from January 1, 2001, through December 31, 2008. Adult long-term survivors of RB were excluded from this analysis.
Within 10 minutes of enucleation of the eye, 3 separate samples of tumor tissue were obtained using methods similar to that described by Shields et al.17 The tumor was then placed at −70°C until submitted for molecular genetic analysis along with the patient's peripheral blood sample. Before 2004, testing was performed at a molecular diagnostic laboratory that provided only DNA sequencing of the RB1 gene (laboratory 1). Subsequently, we used a laboratory that provided sequencing, copy number determination, and RB1 promoter methylation analysis (laboratory 2). The reference laboratory at the hospital used institutional billing for clinical RB1 testing.
Our facilities have had long-standing services, including a cancer genetics clinic (staffed by a medical geneticist and a genetic counselor) that focuses on children and families with hereditary predisposition to cancer, a pediatric neuro-oncology team, pediatric ophthalmologists, and an ocular pathologist at a nearby hospital in the Texas Medical Center. Beginning in the mid-1990s, there was agreement that all children (and long-term survivors) with RB (either bilateral or unilateral) should have a genetic evaluation. The primary referral point for the genetic evaluation occurred at the pediatric oncology clinic. The family was informed of the need for a genetics evaluation, and the cancer genetics counselor was provided the referral information to coordinate the initial appointment. Genetic testing results were disclosed at a follow-up genetics visit that included providing the parents with a genetic counseling letter and copy of the test results. If the proband was found to have a hereditary form of RB, the appropriateness of testing at-risk family members, eg, siblings, for the known familial mutation was discussed in detail. The parents were also provided information about appropriate follow-up for any future children and, where appropriate, were provided a short letter that documented the presence of hereditary RB and the mutation identified so that the parents could distribute this information to other relatives. Parents and at-risk family members were referred back to the ophthalmologist for evaluation and management as indicated.
As part of our RB process, the ocular pathologist ensured that all eyes that were enucleated were adequately handled so that tumor tissue was frozen and stored in the pathology laboratory for further genetic testing (see the “Tumor Analysis” subsection). If subsequently the parents consented to testing, one of the frozen tumor samples was sent along with a blood sample for RB1 analysis. Tumor was included for all patients who had unilateral disease without a family history of RB or, rarely, if the laboratory requested tumor for a bilateral case in which initial testing was uninformative. The ability to have the genetics visit and the discussion of genetic testing after the enucleation substantially simplified the process.
Our process also included a monthly meeting coordinated by the ophthalmologist and attended by the oncologist, pathologist, nurse coordinator, genetic counselor, and geneticist as well as trainees involved in the care of RB patients. At this meeting, all new cases as well as clinical updates on progression of disease and treatment decisions were discussed. This meeting also ensured that all known RB patients had been referred to the genetics clinic. The results of genetic testing were provided, and the genetics team updated the group on the interpretation of the results and the implication for the proband and their family members.
A total of 90 new cases were seen in this period (Figure). Of these, 81 patients were diagnosed with RB, and the remaining 9 children were evaluated because of a family history of RB. The age at RB diagnosis varied from 1 day to 6 years. Quiz Ref IDOf the 81 patients, 34 (42%) had bilateral tumors and 47 (58%) had unilateral tumors. There was a similar rate of success in accomplishing RB1 genetic testing, with 22 of 34 bilateral (65%) and 29 of 47 unilateral (62%) probands having testing completed. Overall, the ability to identify a mutation in the blood sample from patients with bilateral RB was 91% (20 of 22 patients), including those identified through karyotype analysis. As documented in Table 1, the mutation yield increased (from 79% in laboratory 1 to 88% in laboratory 2) when we changed to a DNA diagnostic laboratory that included both RB1 sequencing and copy number analysis of the blood sample (laboratory 2).
The results for the 29 unilateral cases included 17 patients with an informative result (1 or 2 mutations inactivating the RB1 protein product were identified). Again, Quiz Ref IDsensitivity of detecting inactivating events in the RB1 gene was increased (from 57% in laboratory 1 vs 100% in laboratory 2) when we included testing for copy number (number of copies of the exons of the RB1 DNA) and RB1 promoter methylation (addition of methyl groups to the promoter region, which results in silencing of the transcription of the RB1 gene leading to inactivation) (Table 2). Of the 29 unilateral patients undergoing testing, 5 (17%) had a mutation identified in blood samples, documenting a hereditary (germline mutation) form of RB, and 7 patients had fully documented sporadic RB due to 2 somatic events in the tumor tissue and no mutation in the blood. The remaining patients were a mixture of cases in which only 1 mutational event or loss of heterozygosity was identified in the tumor or in which for nonfamilial cases only a blood sample was tested because tumor tissue was not available. In these remaining patients, no mutation was identified in the blood.
Of the 30 RB patients (37%) for whom genetic testing was not accomplished, the most common reason was failure to keep the scheduled genetics appointment despite multiple attempts (16 patients). Other reasons included (1) the cost of the test, despite the fact that the hospital laboratory used institutional billing to increase the likelihood of testing being covered by the patient's health insurance plans; and (2) untimely death of 4 patients (2 with bilateral and 2 with unilateral tumors) from extensive disease or due to complications from therapy. These deaths included patients referred to our center for a clinical treatment trial. Generally, the genetics referral is performed soon after the initial RB diagnosis to obviate this problem.
On the basis of the identification of a germline (germinal) RB1 mutation in a child with RB, we subsequently completed testing of 48 at-risk relatives who were parents and siblings of probands that came to us for initial evaluation (Table 3). Quiz Ref IDSix relatives were found to carry the familial RB1 mutation and required ophthalmologic screening. These 6 individuals were unaffected perhaps due to the fact that they carried mutations typically characterized as having reduced penetrance and expressivity, eg, missense mutations (F755I) or exonic splice changes (V654L). The other 42 relatives were negative for the familial mutation and required no further screening. If genetic testing had not been performed, these 42 individuals would have had to undergo expensive EUA as part of surveillance recommendations. In addition, 9 individuals who did not have RB were evaluated in our clinic because of a positive family history of RB (Figure).
Risks and benefits of testing for genetic susceptibility to cancer depend on the understanding of cancer genetics, the accuracy of the test, and the availability of effective interventions for carriers of the cancer-predisposing genetic mutation.18 Retinoblastoma is a hereditary syndrome for which genetic testing is considered standard of care for at-risk family members.19 Despite these recommendations and the long-term availability of RB1 mutation testing certified by Clinical Laboratory Improvement Amendments, there has been little evaluation of the utility of RB1 mutation analysis in clinical practice. We reviewed our experience with RB1 genetic testing using a multidisciplinary approach. The cohort described here does not appear to represent biased ascertainment because results of evaluating 90 patients demonstrated percentages of bilateral cases (42%), unilateral cases (58%), and positive germline mutation in unilateral cases (17%) similar to percentages reported in other series.6 However, we did note fewer familial cases as only 4 of 81 RB probands had a family history of RB, underscoring that family history of RB should not be a requirement for genetic evaluation. Similar to literature reviews,20,21 the mean age at diagnosis of unilateral RB was 24 months, whereas that of bilateral RB was 15 months.
Most bilateral RB1 cases have a detectable mutation in the blood due to either inheritance from an affected parent or a de novo mutation occurring in the sperm, egg, or embryo. The use of comprehensive methods that include detection of copy number changes or rearrangements increases the likelihood of identifying the genetic basis in the large RB1 gene. However, a negative result may still occur because some mutations may not be identified using current methods (such as those in intronic regions, being large and not sequenced because of difficulty interpreting the result and/or cost; or those in regulatory regions). The reason that neither mutation was identified in the tumor cells in 5 unilateral RB patients is likely due to such abnormalities. In addition, testing of blood can be negative because of mosaicism in the RB patient (the causative mutation is present in only a portion of cells of the body, and sufficient mutation-carrying cells are not present in the blood). Since 2008 when this review concluded, some diagnostic laboratories are using more sensitive tests to detect low-level mosaicism for RB1 mutations in the blood of both bilateral and unilateral RB cases.22,23 Thus, we expect the overall mutation yield to continue to improve.
At our monthly team meeting, the results of genetic testing and their implications for the patient and at-risk relatives were reviewed in detail, facilitating communication across disciplines. The prognosis of RB is highly dependent on prompt diagnosis and evaluation.5 The current focus of management of RB is on optimization of vision while continuing to strive for improved survival. External beam radiotherapy is used almost exclusively in late-diagnosis RB, when all other therapies apart from enucleation have failed. This radiotherapy has been associated with a higher risk of secondary cancers, such as osteosarcoma, in hereditary patients treated for RB.24 There are also concerns about the increased risk of secondary acute myelogenous leukemia after chemotherapy for RB.25 Thus, early detection of RB in at-risk relatives allows for preservation of vision and eye-sparing treatments, with lower morbidity and higher quality of life.
In our study, 17% of patients with unilateral disease were proven to have the hereditary form and thus to be at increased risk for contralateral disease and second malignant neoplasms. However, it is important to note that all unilateral RB patients require surveillance of the contralateral eye with dilated fundus examinations (if older) or EUA (if young), independent of test results, because of the low but real risk of confined somatic mosaicism for an RB1 mutation.6
During this 8-year period, identification of a hereditary mutation provided informative testing for the 48 at-risk family members, of which only 6 were positive and thus required ongoing ophthalmologic screening. Genetic testing results in streamlining of long-term care, in reducing expense, and in alleviating stress from anxiety that the subsequent child will develop RB. Although the cost of RB1 sequencing and deletion/duplication analysis is approximately $1800 for the proband, it is only $340 for the family members when testing for the known familial mutation (as of December 2010). The facility cost associated with 1 EUA at our institution is approximately $1170 ($900 for 45- to 60-minute anesthesia and $270 for the examination). When indirect ophthalmoscopy costs ($670 per eye) and the expense of a 45-minute stay in the postanesthesia care unit ($1138) are added to the EUA, the total expense is approximately $3000 per visit (as of December 2010). Quiz Ref IDAccording to the current surveillance protocols, an at-risk child will require 18 to 26 EUAs (8 EUAs in year 1, 8 EUAs up to age 3, and 6 EUAs up to age 6) with total costs in the first year of life alone of more than $24 000. Thus, the relatively inexpensive, simplified familial mutation testing allows a substantial decrease in the expensive and potentially morbid EUA procedures that would otherwise have been undertaken in these 42 noncarrier individuals. Actual cost estimates are even higher when indirect costs, such as travel and time away from work, as well as morbidity, are taken into account. It should also be noted that prenatal genetic testing cannot be initiated until a causative mutation is identified in the proband, arguing again for the importance of initiating genetic testing in the patient at the time of diagnosis.
Overall, we were able to accomplish genetic analysis in 63% of our patients using a multidisciplinary approach. Integrated care by pediatric oncologists, pediatric ophthalmologists, geneticists, and radiation oncologists with a particular interest in RB is highly recommended.26 Aspects of our process that increased the likelihood of completing genetic evaluation included (1) adequate handling of all enucleated eyes with timely freezing of the tumor samples so that genetic evaluation and consent of parents could be performed after the often difficult period of diagnosis, surgery, and treatment planning; (2) institutional billing by the home institution for testing; (3) the presence of genetics professionals with a focus on cancer genetics (required in many countries in Europe because only a board-certified genetics professional can relay genetic information to patients); (4) a monthly multidisciplinary meeting with a common clinical database that identified patients who were missed by clinic referral; and (5) provision of the test results, including molecular interpretation and implication for the family, to all team members at the monthly meeting.
In conclusion, this study has demonstrated that a multidisciplinary approach has a significant effect on the management of RB. We conclude that (1) coordinated care results in useful genetic evaluations; (2) a range of RB1 mutations were identified in a moderately large population, including novel mutations not previously described; and (3) laboratory testing that includes assessment of promoter methylation and copy number changes improves identification of disease-causing mutations. In summary, use of a multidisciplinary approach ensures that most patients are evaluated in a timely manner and that all members of the team are aware of implications of the results with regard to cancer risk.
Correspondence: Sharon E. Plon, MD, PhD, Feigin Center, Texas Children's Hospital, 1102 Bates St, Room 1200.18, Houston, TX 77030 (email@example.com).
Submitted for Publication: January 6, 2011; final revision received May 25, 2011; accepted May 28, 2011.
Financial Disclosure: None reported.