Improved visual outcome in familial retinoblastoma with late preterm

or early term delivery after prenatal RB1 mutation identification.

Sameh E. Soliman, M.D.1,2,3, Helen Dimaras, Ph.D.2,3,4, Vikas Khetan,

M.B. B.S.5, Elise Héon, M.D., F.R.C.S.C.2,3,4, Helen S.L. Chan, M.B. B.S.,

F.R.C.S.C., Brenda L. Gallie, M.D., F.R.C.S.C.

1Ophthalmology Department, Faculty of Medicine, Alexandria

University

2The Department of Ophthalmology & Vision Sciences, University of

Toronto.

3The Hospital for Sick Children, Toronto, Canada.

4The Division of Visual Sciences, Toronto Western Research Institute,

Toronto, Canada

5Sankara Nethralya Hospital, Chennai, India.

Pediatrics (H.S.L.C.), Molecular Genetics and Medical Biophysics

(B.L.G.), University of Toronto, Toronto, Ontario, Canada; the Division

of Hematology/Oncology (H.D., H.S.L.C.) and Departments of

Ophthalmology & Vision Sciences (E.H., B.L.G., V.K.) and Pediatrics

(H.S.L.C.),; (H.D., B.L.G.);, Toronto, Canada.

Address reprint requests to Dr. Brenda Gallie at the Department of

Ophthalmology and Vision Sciences, the Hospital for Sick Children,

525 University Avenue, Toronto, ON M5G 2L3, Canada, or at

brenda@gallie.ca.

Running Head: early delivery of familial RB.

Key Words: prenatal retinoblastoma, Retinoblastoma gene mutation,

RB1, molecular testing, late pre-term delivery, near-term delivery,

amniocentesis

Submitted as Original Article to JAMA Ophthalmology.

Abstract ( /350)

Importance: The abstract should begin with a sentence or 2 explaining the clinical (or other) importance of the study question.

Objective: State the precise objective or study question addressed in the report (eg, “To determine whether…”). If more than 1 objective is addressed, the main objective should be indicated and only key secondary objectives stated. If an a priori hypothesis was tested, it should be stated.

Objective: To determine overall outcomes of infants with familial retinoblastoma diagnosed

prenatal and delivered preterm compared to infants diagnosed postnatal.

Design: Describe the basic design of the study. State the years of the study and the duration of follow-up. If applicable, include the name of the study (eg, the Framingham Heart Study). As relevant, indicate whether observers were masked to patient groupings, particularly for subjective measurements.

Design: A retrospective, observational study of children born between 1 June 1996 and 1 June 2014

with familial retinoblastoma cared for at SickKids, followed till April 2015.

Setting: Describe the study setting to assist readers to determine the applicability of the report to other circumstances, for example, general community, a primary care or referral center, private or institutional practice, or ambulatory or hospitalized care.

Setting: This study was conducted at the SickKids academic institutional retinoblastoma referral

center.

Participants: State the clinical disorders, important eligibility criteria, and key sociodemographic features of patients. The numbers of participants and how they were selected should be provided (see below), including the number of otherwise eligible individuals who were approached but refused. If matching is used for comparison groups, characteristics that are matched should be specified. In follow-up studies, the proportion of participants who completed the study must be indicated. In intervention studies, the number of patients withdrawn because of adverse effects should be given. For selection procedures, these terms should be used, if appropriate: random sample (where random refers to a formal, randomized selection in which all eligible individuals have a fixed and usually equal

chance of selection); population-based sample; referred sample; consecutive sample; volunteer sample; convenience sample.

Participants: All children with familial retinoblastoma treated at SickKids were included. All children

remain under care at Sickkids.

Intervention(s) for Clinical Trials or Exposure(s) for observational studies: The essential features of any interventions or exposures should be described, including their method and duration of administration. The intervention or exposure should be named by its most common clinical name, and nonproprietary drug names should be used.

Exposure(s): Infants shown on amniocentesis to carry the parent’s RB1 mutant allele were planned

for early pre-term delivery, compared to normal term delivery and postnatal RB1 testing. All children

received treatments for eye tumors.

Main Outcome Measure(s): Indicate the primary study outcome measurement(s) as planned before data collection began. If the manuscript does not report the main planned outcomes of a study, this fact should be stated and the reason indicated. State clearly if the hypothesis being tested was formulated during or after data collection. Explain outcomes or measurements unfamiliar to a general medical readership.

Main Outcome Measures: The hypothesis prior to data collection was that preterm delivery of

infants at near 100% risk of bilateral retinoblastoma safely optimizes visual outcome and minimizes use

of invasive treatments. Primary study outcome measurements were gestational age and ocular location of

first and subsequent tumors; treatments given; eye classification and staging; visual outcome; number of

anesthetics; pregnancy or delivery complications; and estimated overall cost of care.

International Intraocular Retinoblastoma Classification of each eye; laterality; Tumour Node

Metastasis staging; date of last follow-up; and visual outcome.. Data collected included: relation to

proband; sex; gestational age at birth;; date of RB1 molecular testing, type of sample tested and result;

Results: The main outcomes of the study should be reported and quantified, including baseline characteristics and final included/analyzed sample. Include absolute numbers and measures of absolute risks (such as increase/decrease or absolute differences between groups), along with confidence intervals (for example, 95%) or Pvalues. Approaches such as number needed to treat to achieve a unit of benefit may be included when appropriate. Measures of relative risk also may be reported (eg, relative risk, hazard ratios) and should include confidence intervals. Studies of screening and diagnostic tests should report sensitivity, specificity, and likelihood ratio. If predictive value or accuracy is reported, prevalence or pretest likelihood should be given as well. All randomized clinical trials should

include the results of intention-to-treat analysis, and all surveys should include response rates.

Results: Of 21 infants shown to carry their parent’s RB1 mutation, 12 had been tested prenatally and

9 after birth. Of the infants tested prenatally, 9 were induced at 36-38 weeks gestation and 3 were born

spontaneously preterm. Immediate postnatal examination revealed vision-threatening tumors present only

in 25% (3/12) of infants prenatally diagnosed, compared to 67% (6/9) of those diagnosed postnatally. All

patients eventually developed tumors in both eyes. Good vision was maintained in all prenatally

diagnosed patients; treatments included focal therapy (all); later systemic chemotherapy (5), enucleation

and stereotactic radiation (1). Full-term infants received focal therapy (all), systemic chemotherapy (4),

stereotactic radiation (2), and enucleation of one eye (4), with poorer visual outcome.

Conclusions and Relevance: Provide only conclusions of the study that are directly supported by the results, along with implications for clinical practice or health policy, avoiding speculation and overgeneralization. Indicate whether additional study is required before the information should be used in usual clinical settings. Give equal emphasis to positive and negative findings of equal scientific merit. Also, provide a statement of relevance indicating implications for clinical practice or health policy, avoiding speculation and overgeneralization. The relevance statement may also indicate whether additional study is required before the information should be used in clinical settings.

Conclusions and Relevance: Prenatal diagnosis of retinoblastoma followed by late

preterm and near-term delivery facilitated expedient intervention and optimal outcomes.

Retinoblastoma, the most common primary ocular malignancy in children, is initiated when both

alleles of the RB1 tumor suppressor gene are inactivated in a precursor retinal cell, and progresses when

mutations in other specific genes occur.1,2 Both alleles may be lost only in the somatic cell from which the

tumor arises, however, in about 50% of children, a germline mutation predisposes to the development of

multiple retinal tumors during childhood, and other cancers later in life. Ten percent of patients display a

family history of disease, inheriting a family-specific mutation from a parent.1,3

Children with RB1 germline alleles may already have retinoblastoma tumor(s) at birth, which are

often in the posterior pole of the eye where they threaten vision.4-8 Preservation of vision with treatment of

these small tumors is often difficult, because focal treatment in proximity to the optic nerve and macula

may damage vision. Most of these children will develop more tumors in the first year of life, which tend

to be located peripherally. The child is bilaterally affected in either simultaneous or sequential

involvement. 4,7 Low penetrance mutations (10% of familys)3 and mosiacism result in fewer tumors and

more unilaterally affected children9. The timing of first tumors after birth has not been studied.

It is recommended that infants with a family history of retinoblastoma be screened as soon as

possible after birth and repeatedly for the first few years of life, including under anaesthesia, aiming at

early diagnosis when tumors are small and easy to treat with ocular and visual salvage.6,7,10

Preterm birth is defined as a live birth occurring before completion of 37 weeks gestation.

Subcategories of preterm birth include: extremely preterm (<28 weeks gestation), very preterm (28 to <32

weeks) and moderate-to-late preterm (32 to <37 weeks). Full term birth is generally defined as a live birth

occurring at 40 weeks gestation. Infants born after completion of 37 and before 39 weeks gestation are

technically considered early term. (8-9). 11,12 The main concern with late preterm or early term delivery is

its reported effect on neurological and cognitive development and later school performance ,13-15 but visual

dysfunction from a larger macular tumor can cause similar neurocognitive defects due to blindness16

despite never studied in a comparative manner.

We now present the first report of prenatal genetic screening and late preterm or early term delivery

for treatment of retinoblastoma for children demonstrated to carry the RB1 mutant allele of a parent. We

show that for children at 50% risk to inherit a germline RB1 mutant allele, prenatal molecular diagnosis

and preterm delivery allowed detection and treatment of small, early tumors, resulting in lower treatment

morbidity, better tumor control and visual outcome, compared to children born full term at 39-40 weeks.

Methods

Study DesignResearch ethics board approval (REB approval number 1000028725) was obtained from The

Hospital for Sick Children (SickKids) for a retrospective review of medical records of all children with

familial retinoblastoma seen at SickKids, born between 1 June 1996 and 1 June 2014. Data collected

included: relation to proband; laterality of retinoblastoma in proband; sex; gestational age at birth;

pregnancy, prenatal abdominal ultrasound if done; delivery or perinatal complications; type of genetic

sample tested and result; penetrance of RB1 mutaion; age and location of first and all subsequent tumor

(s) in each eye; treatments used; number of anaesthetics; International Intraocular Retinoblastoma

Classification17 of each eye (IIRC); Tumor Node Metastasis (TNM) staging for eyes and child18; treatment

duration; date of last follow-up; and visual outcome at last follow-up in Snellen and LogMAR values.

RB1 mutation testing was performed by Retinoblastoma Solutions before 2013, and Impact Genetics after

2013, as previously described.19

The corrected age for gestation for each child was calculated (taking 39 weeks as full term). Vision

threatening tumors were defined as in close to optic nerve or macular area (IIRC17 Group B or worse.

Treatment burden was evaluated by an impact score (Figure 1) based on i) duration of active treatment

(time from diagnosis to last treatment), ii) use of systemic chemotherapy or radiation, and iii) number of

examinations under anesthesia (EUAs). Treatment success was defined as avoidance of enucleation or

external beam irradiation. Good visual outcome was defined as visual acuity > 20/200 (>0 in 1-LogMAR

scale) (cut edge of legal blindness). A blind child is defined as best eye visual acuity < 20/200 (<0 in 1-

LogMAR scale).

StatisticsBasic descriptive statistics (student t-test, chi square test and Fisher exact test) were used for

comparisons between patients who underwent prenatal testing and preterm delivery (Cohort 1) and those

who were diagnosed post-natal (Cohort 2).

Results

Patient DemographicsThe records of 21 familial retinoblastoma children were reviewed (11 males, 10 females)

(Supplementary Table 1). Diagnosis was by observation of prenatal retinoblastoma tumor (child #9) or

postnatal tumor (child #8) or postnatal testing for the parental RB1 mutation for Cohort 1: 6 were

delivered full term and 3 late preterm because of pregnancy-induced hypertension (#7), fetal ultrasound

evidence of retinoblastoma20 (#9) or spontaneous (#8). Twelve children (57%) (Cohort 2) were prenatally

diagnosed to carry an RB1 mutation and planned for late preterm or early term delivery: 3 were

spontaneously premature (#10, 13, 15; 28-37 weeks gestation) and 9 were referred to a high-risk

pregnancy unit for elective late preterm or early term delivery (36-38 weeks gestation).

Molecular diagnosisAll study subjects were offspring of retinoblastoma probands. Nineteen probands were bilaterally,

and 2 were unilaterally affected (mother #8, father #19). The familial RB1 mutations were previously

detected except for the unilaterally affected parent of #8, who was never testsed and understood that her

children had no risk since she was unilaterally affected. whose mutation was identified after finding her

son’s mutation as she was tested previously and considered to be a non germline mutation. Cohort 1

children (#1-9) were tested postnatal on blood; Cohort 2 children (#10-21) were tested prenatal on

amniocentesis at 16-33 weeks gestation.

Null RB1 mutations were present in 16 families; 5 had low penetrance RB1 mutations (whole gene

deletion #19; weak splice site mutations #15, 18, 21; and C712R19). No proband in this study was mosaic

for the RB1 mutation. All study subjects were eventually bilaterally affected. No tumors were detected at

birth (IIRC17 Group 0) in 7/15 (47%) infants and 17/30 (57%) eyes with null RB1 mutations, and in 5/5

(100%) infants and 10/10 (100%) eyes with low penetrance mutations (p=0.04* for patients, p=0.02* for

eyes; Fisher exact test) (Table 1).

The age at first tumor in either eye was significantly younger for those with null mutations (mean

89, median 74 days), than those with low penetrance mutations (mean 134, median 119 days) (P=0.04*,

Mann Whitney test). However, the gestational age at first tumor for those with null mutations (mean 77,

median 43 days) was not significantly different than for those with low penetrance mutations (mean 110,

median 81 days) (P=0.24*, Mann Whitney test). (Child #8 was excluded from these calculations as the

child was first examined at 3 months of age with Group A/D tumors, so age at first diagnosis is

unknown.)

Stage of Tumors at Birth

Thirty-three percent (3/9) of Cohort 1 and 75% (9/12) of Cohort 2 were free of visible tumor at birth

(Table 1a, Figure 1) (p=0.09). We assumed that child #8 had tumor at birth since he had a group D IIRC17

eye at 3 months of age. Of eyes, 79% (19/24) of Cohort 1 eyes were tumor-free at birth, compared to 33%

(6/18) of Cohort 2 eyes (p=0.026*, Chi Square test), excluding the IIRC17 Group A eye of child #8 (Table

1b).

All patients eventually developed tumors in both eyes regardless of whether their RB1 mutation was

full or low penetrance. Tumors emerged first in the macular and peri-macular region (IIRC17 Group B), as

previously described21. The median gestational age of diagnosis of 24 IIRC17 A eyes (< 3mm and away

from optic nerve and fovea) was 103 days, and of 16 IIRC17 B eyes (all threatening optic nerve and fovea,

6 also >3 mm) was 38 days, significantly younger reflecting the early development of visually threatening

tumors (Table 3?). Bilateral IIRC17 Group A eyes were present at initial diagnosis (optimal situation for

achieving good vision with minimally invasive therapy) in 2/9 (22%) children in Cohort 1 and 7/12 (58%)

in Cohort 2 (Table 2a). IIRC17 Group A was the initial diagnosis of 9/18 (50%) eyes in Cohort 1, and

11/22 (68%) eyes in Cohort 2 (p=0.15, Table 2b).

Treatment Course All infants were frequently examined from birth onwards (except child #8 who presented at age 3

months) as per the National Retinoblastoma Strategy Guidelines for Care.10 Cohort 1 patients were treated

with focal therapy (all), chemotherapy (4), stereotactic radiation (2), and enucleation of one eye (4)

(Supplementary Table 1, Figures 1, 3). Cohort 2 patients were treated with focal therapy (all); later

systemic chemotherapy (5), enucleation of one eye and stereotactic radiation (1) (Figure 1, 2). Successful

treatment by focal therapy alone (avoidance of systemic chemotherapy or EBRT) was possible in 2/9

(22%) of Cohort 1 and 7/12 (58%) of Cohort 2 (PVALUE?) (Table 3, Figure 2).

The treatment burden showed no statistical significant difference between Cohort 1 and 2

(percentage of patients requiring systemic therapy and active treatment duration). The mean active

treatment duration was 579 days (0-2101 days) in Cohort 1 compared to 473 days (0-971 days) in Cohort

2.

REPLACE WITH (TREATMENT?) IMPACT SCORE……..AS DEFINED…….ABOVE

OutcomesThere were no adverse effects associated with induced or natural preterm or early term birth, and no

pregnancy, delivery or perinatal complications reported for any of the infants. Follow up (years) was

overall mean 8 (median 5.6), Cohort 1 mean x (median y), and Cohort 2 mean x years (median y)

(Supplementary Table 1).

Neither enucleation or external beam irradiation were required (defined as success) in 44% of

Cohort 1 and 92% of Cohort 2 (P=0.046*, Fisher exact test). Kaplan Meier ocular survival for Cohort 1

was 67% compared to 92% for Cohort 2 (statistics?) (Figure x). No child developed extra ocular or

metastatic disease; all are alive and well at last followup.

Visual outcomes were good for 56% of eyes in Cohort 1 and 92% of eyes in Cohort 2 (P=0.010*,

Fisher exact test). Children were legally blind (visual acuity less than 20/200 both eyes) in 22% of Cohort

1 and 0% of Cohort 2 (p=0.174, Fisher exact test)

Treatment success (avoidance of enucleation and/or stereotactic radiation) and good vision per eye

was documented 50% (9/18) of Cohort 1 and 88% (21/24) of Cohort 2 (p=0.014*, Fisher exact test)

(Table xb, Figure 1). A negative correlation was found between gestational age and final visual outcome

(r=-0.24) with better visual outcome in earlier deliveries. (Figure )

Discussion

In the first study of its kind, we report that prenatal molecular diagnosis of familial retinoblastoma

and elective late-preterm or early term delivery allowed monitoring for, and treatment of, tumors as they

emerged, which resulted in better ocular and visual outcomes and less severe medical interventions in

very young children. This data illustrates that for infants with close to 100% risk of retinoblastoma in

both eyes because they carry an RB1 mutant allele, the risk of vision and eye loss despite intensive

therapies, outweighs the risks associated with induced late preterm delivery (Figure 1). Consistent with

previous reports,{Abramson, 1998 #22158} 62% of children with a germline gene mutation already had

tumors at full term birth. This reduced to 31% when the germline mutation is prenatally detected and

earlier delivery (late preterm or early term) was planned.

It is practical to identify 96% of the germline mutations in bilaterally affected probands and to

identify the >15% of unilateral probands who carry a germline gene mutation.{Rushlow, 2009

#22181;Lohmann, 2000 #22176} When the family's unique mutation is identified in the proband,

molecular testing of family members can determine who else carries the mutation and is at risk to develop

retinoblastoma. We report on 12 infants identified by in utero molecular testing to carry the mutant RB1 allele of a parent. The 50% of tested infants who did not inherit their family’s mutation require no

surveillance, can be born at full term and do not need examinations to detect tumors, since they are at no

greater risk of developing retinoblastoma than the general population.

Without molecular information, repeated retinal examination is recommended for all first degree

relatives until age 7 years, the first 3 years under general anesthesia.10 Multiple studies now suggest

deleterious effects of multiple general anesthetics in early infancy on the neurocognitive development of

the child.22-24 Such repeated clinical screening also imposes psychological and financial burden on the

children and families. Identification by early molecular RB1 testing of the children who are not at risk

and require no clinical intervention cost significantly less than direct costs than clinical screening for

tumors.19,25

Optimal treatment for retinoblastoma includes combined therapeutic modalities to optimize vision

and minimize treatment morbidity, while achieving tumor control. Our study shows that the burden,

blah …. are less when tumors are identified at younger age and before full term, as previously

suggested.26 However, retinoblastoma treatment in the first 3 months of life is a challenge since these

young children may not have sufficient renal function for full dose systemic therapies. In our study child

#9, who had a tumor at 36 weeks gestation large enough to detect by obstetrical ultrasound, showed drug-

resistant tumor following reduced-dose chemotherapy, ultimately requiring enucleation of one eye. The

only treatment options at this age are focal therapy (laser and cryotherapy) and periocular chemotherapy.27

The earliest tumors commonly involve the macular or paramacular region, dangerously risking loss

of central vision, while tumors that develop later on are usually peripheral, where they have less visual

impact.5,27-30 In our cohort, the risk of having a vision threatening tumor dropped from 39% to 17% by

prenatal mutation detection and planned earlier delivery. TALK ABOUT THE BIOLOGICAL

MESSAGE SHOWN IN THE NEW GRAPH (COMING SOON). AND WE NEED THE DATA ON

ALL TUMORS THEY GOT IN A AND B EYES, (CAN’T COUNT NEW VS SECONDARY IN C

AND D EYES)

Macular and paramacular tumors are difficult to manage by laser therapy or application of a

radioactive plaque, since these will damage the optic nerve or central vision. Systemic chemotherapy

effectively shrinks tumors such that focal therapy can be applied with minimal visual damage. In our

setting, the Toronto Protocol using high dose, short duration cyclosporine to counteract multidrug

resistance31,32 has allowed many retinoblastoma tumors to be treated with combination chemotherapy and

focal therapy without resorting to radiation. In our experience, infants as young as 30 days tolerate the

Toronto Protocol with cyclosporine A, 31,32 however systemic chemotherapy in neonates has other

associated morbidities. We recognize the conventional recommendation to reduce chemotherapy dosages

by 50%, particularly for infants in the first three months of life, to offset the immaturity of their liver and

kidneys,33-35 but note that this also breeds the optimal conditions for cancer cells to develop multidrug

resistance, making later recurrence difficult to treat. The development of periocular topotecan for

treatment of small-volume retinoblastoma36 also assisted in the number of patients that were able to be

treated by focal therapy alone, avoiding systemic modalities on the young infants, and a greater rate of

eye salvage with good visual outcome (Table 2).

Imhof et al7 in the Netherlands screened 135 children at risk of familial retinoblastoma 1-2 weeks

after birth without molecular diagnosis and discovered 17 cases of familial RB (13% of screened children

at risk) and 70% of them had RB in at least one eye at first examination and 41% of eyes had vision

threatening tumor to the macula. 41% (7/17) of patients had failure of treatment (EBRT or enucleation)

and one case of metastasis. 73.5% of eyes (27/34) had good visual acuity (defined by vision >20/100) that

will reduce to 56% (19/27) if we consider eyes with EBRT as failure. These results correspond to our

postnatal screening cohort showing similar results. On the contrary, the prenatal diagnosis and planned

earlier delivery cohort showed less vision threatening tumors (17%), less treatment failure (8%) and better

visual outcome (88%).

Early screening of at risk infants with positive family history as soon an possible after birth is the

internationally accepted model (whether intensive screening is utilized or not).7,37 Here we propose the

prenatal screening of the known mutation in the probands by amniocentesis in the second half of

pregnancy where the risks of miscarriage are minimal (0.1-1.4%).38,39 For those who are confirmed to

have the mutation; planned late preterm or early term delivery at 36-38 weeks of gestation and as a result

a smaller tumor with less macular involvement leading to better visual outcome is anticipated. there was

no difference between the two Cohorts in the treatment burden and the systemic chemotherapy usage as

we didn't change the treatment course by early delivery but changed the treatment outcome by catching

the tumors at earlier stage also multiple focal treatments in both Cohorts were for small new tumors that

occurred due to the nature of the germline tumor and not related to early delivery or prenatal detection.

The main concern with late preterm or early term delivery is its reported effect on neurological and

cognitive development and later school performance (30-32),13-15 but visual dysfunction from a larger

macular tumor can cause similar neurocognitive defects due to blindness16 despite never studied in a

comparative manner. So, earlier delivery must be discussed thoroughly through the team of neonatologist

ophthalmologist and oncologist to reach the best timing for better outcome26 so rather than focusing on

the combination of treatments to tackle burdensome disease, we showed safe preterm delivery resulted in

a decreased tumor burden at birth that was significantly easier to treat (Figure 2, Table 2). Safe preterm

delivery resulted in more infants born tumor-free, facilitating frequent surveillance to detect tumors as

they emerged, and focal therapy of smaller, easier to control masses, causing minimal damage to vision

(Figure 1,2).

Counseling on reproductive risks is imperative for families affected by retinoblastoma even in

unilateral probands. In developed countries; where current therapies result in extremely low mortality,

most retinoblastoma patients will survive to have children. Prenatal diagnosis in the published literature

has been cited as useful in preimplantation genetics (to ensure an unaffected child) or to inform parents

who wish to terminate an affected pregnancy40. There have been two prior reports indicating pre-natal

molecular testing for retinoblastoma; in one, the fetus sibling of a proband was found not to carry the

sibling’s mutation41, and in the other, 3 of 5 tested fetuses of a proband were terminated once molecular

testing confirmed the mutation in the offspring.42 We are first to report that elective safe late-preterm

delivery of prenatally diagnosed infants with retinoblastoma results in improved outcomes. It is our

experience that for retinoblastoma survivors and their relatives who understand fully the underlying risks,

they are more interested in early diagnosis to optimize options for therapy in affected babies rather than to

consider termination of pregnancy. We also surmise that since germline mutations predispose to future,

second cancers in affected individuals, perhaps it is worth investigating the role of cord blood banking

infants that are prenatally molecularly diagnosed with retinoblastoma. A long-term study could show the

impact of such an approach to patient outcomes in their adulthood. We conclude that since infants with

familial retinoblastoma are likely to develop vision-threatening macular tumors, prenatal molecular

diagnosis and safe, late-preterm delivery will increase the chance of good visual outcome with decreased

treatment associated morbidity.

References

1. Corson TW, Gallie BL. One hit, two hits, three hits, more? Genomic changes in the development of retinoblastoma. Genes Chromosomes Cancer. 2007;46(7):617-634.

2. Dimaras H, Khetan V, Halliday W, et al. Loss of RB1 induces non-proliferative retinoma: increasing genomic instability correlates with progression to retinoblastoma. Hum Mol Genet. 2008;17(10):1363-1372.

3. Lohmann DR, Gallie BL. Retinoblastoma. In: Pagon RA, Adam MP, Ardinger HH, et al., eds. GeneReviews(R). Seattle (WA)2000.

4. Abramson DH, Du TT, Beaverson KL. (Neonatal) retinoblastoma in the first month of life. Archives of ophthalmology. 2002;120(6):738-742.

5. Abramson DH, Mendelsohn ME, Servodidio CA, Tretter T, Gombos DS. Familial retinoblastoma: where and when? Acta Ophthalmol Scand. 1998;76(3):334-338.

6. Noorani HZ, Khan HN, Gallie BL, Detsky AS. Cost comparison of molecular versus conventional screening of relatives at risk for retinoblastoma. American journal of human genetics. 1996;59(2):301-307.

7. Imhof SM, Moll AC, Schouten-van Meeteren AY. Stage of presentation and visual outcome of patients screened for familial retinoblastoma: nationwide registration in the Netherlands. Br J Ophthalmol. 2006;90(7):875-878.

8. Abouzeid H, Schorderet DF, Balmer A, Munier FL. Germline mutations in retinoma patients: relevance to low-penetrance and low-expressivity molecular basis. Molecular vision. 2009;15:771-777.

9. Rushlow D, Piovesan B, Zhang K, et al. Detection of mosaic RB1 mutations in families with retinoblastoma. Human mutation. 2009;30(5):842-851.

10. National Retinoblastoma Strategy Canadian Guidelines for Care / Stratégie thérapeutique du rétinoblastome guide clinique canadien. Canadian journal of ophthalmology. 2009;44(Supp 2):S1-88.

11. ACOG Committee Opinion No 579: Definition of term pregnancy. Obstet Gynecol. 2013;122(5):1139-1140.

12. Born Too Soon: The Global Action Report on Preterm Birth. Geneva: World Health Organization;2012.

13. Cheong JL, Doyle LW. Increasing rates of prematurity and epidemiology of late preterm birth. Journal of paediatrics and child health. 2012;48(9):784-788.

14. Woythaler MA, McCormick MC, Smith VC. Late preterm infants have worse 24-month neurodevelopmental outcomes than term infants. Pediatrics. 2011;127(3):e622-629.

15. Poulsen G, Wolke D, Kurinczuk JJ, et al. Gestational age and cognitive ability in early childhood: a population-based cohort study. Paediatr Perinat Epidemiol. 2013;27(4):371-379.

16. Bedny M, Saxe R. Insights into the origins of knowledge from the cognitive neuroscience of blindness. Cogn Neuropsychol. 2012;29(1-2):56-84.

17. Murphree AL. Intraocular retinoblastoma: the case for a new group classification. Ophthalmology clinics of North America. 2005;18:41-53.

18. The 7th edition AJCC staging system for eye cancer: an international language for ophthalmic oncology. Archives of pathology & laboratory medicine. 2009;133(8):1197-1198.

19. Richter S, Vandezande K, Chen N, et al. Sensitive and efficient detection of RB1 gene mutations enhances care for families with retinoblastoma. American journal of human genetics. 2003;72(2):253-269.

20. Sahgal A, Millar BA, Michaels H, et al. Focal stereotactic external beam radiotherapy as a vision-sparing method for the treatment of peripapillary and perimacular retinoblastoma: preliminary results. Clin Oncol (R Coll Radiol). 2006;18(8):628-634.

21. Munier FL, Balmer A, van Melle G, Gailloud C. Radial asymmetry in the topography of retinoblastoma. Clues to the cell of origin. Ophthalmic genetics. 1994;15(3-4):101-106.

22. DiMaggio C, Sun LS, Ing C, Li G. Pediatric anesthesia and neurodevelopmental impairments: a Bayesian meta-analysis. Journal of neurosurgical anesthesiology. 2012;24(4):376-381.

23. Ing C, DiMaggio C, Whitehouse A, et al. Long-term differences in language and cognitive function after childhood exposure to anesthesia. Pediatrics. 2012;130(3):e476-485.

24. Wilder RT, Flick RP, Sprung J, et al. Early exposure to anesthesia and learning disabilities in a population-based birth cohort. Anesthesiology. 2009;110(4):796-804.

25. Houdayer C, Gauthier-Villars M, Lauge A, et al. Comprehensive screening for constitutional RB1 mutations by DHPLC and QMPSF. Human mutation. 2004;23(2):193-202.

26. Dimaras H, Kimani K, Dimba EA, et al. Retinoblastoma. Lancet. 2012;379(9824):1436-1446.27. Gombos DS. Retinoblastoma in the perinatal and neonatal child. Semin Fetal Neonatal Med.

2012;17(4):239-242.28. Abramson DH, Gamell LS, Ellsworth RM, et al. Unilateral retinoblastoma: New intraocular

tumours after treatment. Br.J.Ophthalmol. 1994;78(9):698-701.29. Abramson DH, Greenfield DS, Ellsworth RM. Bilateral retinoblastoma. Correlations between age

at diagnosis and time course for new intraocular tumors. Ophthalmic paediatrics and genetics. 1992;13(1):1-7.

30. Abramson DH, Servodidio CA. Retinoblastoma in the first year of life. Ophthalmic paediatrics and genetics. 1992;13(4):191-203.

31. Salmonsen PC, Ellsworth RM, Kitchen FD. The occurrence of new retinoblastoma after treatment. Ophthalmology. 1979;86:840-843.

32. Gallie BL, Budning A, DeBoer G, et al. Chemotherapy with focal therapy can cure intraocular retinoblastoma without radiotherapy. Archives of ophthalmology. 1996;114(11):1321-1328.

33. Chan HS, DeBoer G, Thiessen JJ, et al. Combining cyclosporin with chemotherapy controls intraocular retinoblastoma without requiring radiation. Clin Cancer Res. 1996;2(9):1499-1508.

34. Campbell AN, Chan HS, O'Brien A, Smith CR, Becker LE. Malignant tumours in the neonate. Archives of disease in childhood. 1987;62(1):19-23.

35. Crom WR, Glynn-Barnhart AM, Rodman JH, et al. Pharmacokinetics of anticancer drugs in children. Clin Pharmacokinet. 1987;12(3):168-213.

36. Mallipatna AC, Dimaras H, Chan HS, Heon E, Gallie BL. Periocular topotecan for intraocular retinoblastoma. Archives of ophthalmology. 2011;129(6):738-745.

37. Rothschild PR, Levy D, Savignoni A, et al. Familial retinoblastoma: fundus screening schedule impact and guideline proposal. A retrospective study. Eye (Lond). 2011;25(12):1555-1561.

38. Akolekar R, Beta J, Picciarelli G, Ogilvie C, D'Antonio F. Procedure-related risk of miscarriage following amniocentesis and chorionic villus sampling: a systematic review and meta-analysis. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology. 2015;45(1):16-26.

39. Tabor A, Vestergaard CH, Lidegaard O. Fetal loss rate after chorionic villus sampling and amniocentesis: an 11-year national registry study. Ultrasound in obstetrics & gynecology : the official journal of the International Society of Ultrasound in Obstetrics and Gynecology. 2009;34(1):19-24.

40. Dommering CJ, Garvelink MM, Moll AC, et al. Reproductive behavior of individuals with increased risk of having a child with retinoblastoma. Clin Genet. 2012;81(3):216-223.

41. Lau CS, Choy KW, Fan DS, et al. Prenatal screening for retinoblastoma in Hong Kong. Hong Kong Med J. 2008;14(5):391-394.

42. Castera L, Gauthier-Villars M, Dehainault C, et al. Mosaicism in clinical practice exemplified by prenatal diagnosis in retinoblastoma. Prenatal diagnosis. 2011;31(11):1106-1108.

Figures

Tables