Ocular Manifestations in Children Born After In Vitro Fertilization FREE

THE NUMBER of infants born after in vitro fertilization (IVF) is increasing steadily as the techniques of micromanipulation and neonatal intensive care evolve. The prevalence of neonatal and postnatal developmental abnormalities among this fast-growing population of children remains, however, largely unknown. Few prospective studies on this subject have been performed. It has been reported that the rate of malformations detected in children born after IVF is similar to that observed in the non-IVF newborn population.^{1 - 6} Some studies,^{7 - 9} however, disclosed an elevated rate of chromosomal aberrations among the IVF-born offspring. These chromosomal abnormalities affected primarily sex chromosomes.

The purpose of the present study is to report the ocular findings (and abnormalities) detected in 47 children born after IVF and evaluated in our pediatric ophthalmology unit between July 1, 1994, and December 31, 1999. The possible association of these ocular abnormalities with the intracytoplasmic sperm injection (ICSI) method of micromanipulation is also examined.

Forty-seven children (25 girls and 22 boys) born after IVF were examined in the Pediatric Ophthalmology Unit at Hadassah University Hospital in Jerusalem, Israel, between July 1, 1994, and December 31, 1999. Twenty children (43%) were referred by their treating pediatrician, 18 (38%) were examined at the request of the parents because of a heightened degree of anxiety, and 9 (19%) were referred to our clinic for a second opinion because of the presence of ocular abnormalities.

An attempt at careful and complete obstetric history taking was performed in all cases. This included the presence or absence of previous pregnancies; length of infertility period and its cause; ingestion of drugs before, during, and after IVF; and type of IVF performed. Missing clinical obstetric and gynecologic details were obtained in some cases through a telephone questionnaire with the mothers.

All children underwent a thorough ophthalmologic examination. In preverbal children, assessment of the visual acuity was performed as previously described.¹⁰ In short, fixation behavior was assessed using the smallest age-appropriate target. Monocular fixation was recorded for 3 separate factors: (1) quality and accuracy (good, fair, poor), (2) location (central vs eccentric), and (3) duration (maintained vs sporadic). For children with poor fixation, the visual abilities were evaluated using an optokinetic drum. In verbal children, this was based on identifying familiar pictures and shapes either by single or linear optotypes. In older and cooperative children, the Tumbling E or Snellen charts was used.

Ocular motility examination, including assessment of ductions and versions in all cardinal positions of gaze, was also performed. In the very young or inattentive children, the doll's head movement was used to rule out limitation of movements. Strabismus was assessed by cover-uncover and alternate-cover tests for near in all children and also for distance in older cooperative ones. In cooperative children older than 3 years, binocular functions were evaluated using the Titmus Fly Stereotest, whereas fusional capacity was assessed by the Worth 4-dot test.

Survey of the external structures and anterior segment was carried out at the slitlamp examination. Lens anatomy and anterior vitreous were reexamined after maximal pupil dilation.

Cycloplegic refraction was carried out 30 minutes after instillation of 1% cyclopentolate hydrochloride (Cyclogyl) and 0.5% tropicamide (Mydramide) delivered twice at 10-minute intervals.

Fundus was examined by indirect ophthalmoscopy with special emphasis on optic nerve anatomy and color, macula and foveal reflexes, and configuration of retinal blood vessels in the posterior pole. Detailed examination of the periphery was carried out only in eyes with unclear findings in the posterior pole.

Data are given as mean ± SD.

The 47 children referred to our unit were the siblings of 68 children born after 41 IVF-induced pregnancies. The remaining 21 siblings were reported by the parents to have no ocular abnormality and therefore were not brought to our unit. Thirty (64%) of the 47 children were part of a twin (27 or 57%) or triplet (3 or 6%) pregnancy. The birth weight of the entire group was 2335 ± 817 g (range, 924-4300 g). For the single fetus birth, the birth weight was 2921 ± 922 g (range, 924-4300 g), and for multiple fetuses birth, the single birth weight was 1990 ± 559 g (range, 1080-3470 g). The birth weight was significantly lower for the multiple fetus pregnancies(P = .002) (Table 1).

The gestation age for the entire group was 35.0 ± 3.7 weeks (range, 28-40 weeks; median, 34 weeks). For the 9 single fetus pregnancies, the gestational age was 37.7 ± 3.5 weeks (range, 29-40 weeks; median, 40 weeks). The gestational age in multiple fetus pregnancies was significantly lower (33.8 ± 3.1 weeks; range, 28-40 weeks; median, 33 weeks; P = .001) (Table 1).

Despite efforts, including telephone questionnaires, thorough and detailed obstetric data regarding the type of IVF technique and pregnancy were only obtained in 33 (80%) of the 41 pregnancies. The age of the mothers at time of conception was 32.1 ± 4.7 years (range, 23-43 years). The fathers' age was 34.8 ± 6.5 years (range, 26-51 years) (Table 2).

The period of known infertility was 4.9 ± 4.6 years (range, 0-17 years; median, 3.5 years). Infertility was attributed to a combination of causes in 8 of the women. Sperm quality and/or oligospermia were the probable causes for infertility in 15 of 33 women for which reliable information was available. Ovulation-related problems were implicated in 12 of 33 cases, and mechanical-related abnormalities of the female reproductive system were the cause of infertility in 11 of 33 cases. In 4 cases, the cause for the infertility could not be ascertained.

The IVF technique was supplemented by ICSI in 16 cases (48%). Sperm donations were used for 4 of the pregnancies. No intrauterine fetal reduction was performed for any of the pregnancies in the study group.

Delivery was vaginal in 16 (47%) of 34 women, whereas a cesarean section was performed in 18 (53%).

The children were first examined in our clinic at an age of 2 months to 5 years (mean, 2.3 ± 5.8 years; median, 9.6 months). At the time of examination, 12 children were verbal and 35 children were not. Of 70 eyes of nonverbal children, 60 (86%) were classified as having vision "normal for age", 4 (6%) had "fair" or "questionably normal", and 6 (9%) had "poor" or"definitely abnormal" vision. In the 12 verbal children, visual acuity was 6/10 or better in 17 eyes, 6/12 to 6/30 in 2 eyes, 6/36 to 6/90 in 1 eye, 6/120 to light perception in 1 eye, and no light perception in 3 eyes.

Reliable cycloplegic refraction was achieved in 84 eyes. In 5 eyes no retinoscopic reflex was obtained because of advanced intraocular pathologic conditions, and in 5 eyes accurate refraction was not possible. Two eyes were aphakic at the time of examination because they had undergone cataract surgery without intraocular lens implantation elsewhere. The cycloplegic refraction of the remaining 82 fully assessable eyes disclosed an emmetropia (spherical equivalent refractive error of 0.00 to +1.00 diopters [D]) in 22 eyes (27%). A hypermetropic refraction of more than +1.0 D was found in 47 eyes (57%), and a myopic refraction (spherical equivalent refractive error of –0.50 D or more) was found in 13 (16%) eyes. In 20 eyes, the refractive error was larger than ±3.0 D. Eight children had an anisometropia of more than 1.0 D.

In 43 of the children, a full assessment of ocular motility and cover tests was achieved. Thirty-one were orthophoric (72%), 7 (16%) were esotropic, and 5 (12%) were exotropic.

Major ocular malformations were observed in 12 (26%) of 47 children, affecting 19 eyes (Table 3). The conditions included Coats disease (2 eyes of 2 children), congenital cataract(4 eyes of 2 children), coloboma of uvea (3 eyes of 2 children), hypoplastic optic nerve heads (3 eyes of 2 children), frank idiopathic optic atrophy (4 eyes of 2 children), congenital glaucoma (2 eyes of 1 child), and retinoblastoma(1 eye in 1 child).

Retinopathy of prematurity (ROP) was observed in 2 children. One child with a gestational age of 29 weeks and a birth weight of 924 g reached ROP stage II in both eyes that regressed spontaneously. One child, part of a triplet(gestational age, 30 weeks; birth weight, 2060 g), was diagnosed as having ROP stage V. Despite cryotherapy and several retinal surgical procedures, visual function in both eyes of this child remained very poor.

Additional, sporadic pathologic conditions were observed in this group of children. These included periocular hemangiomas in 3 children and Duane retraction syndrome in 1 child.

Major ocular abnormalities were observed in 12 of the 47 children born after IVF pregnancies. This 26% incidence of ocular pathologic conditions in our cohort is higher than expected. Obviously, a certain referral bias to our clinic and a heightened level of parent anxiety are most probably important factors that affect these data, with an imbalance influence on this surprisingly high number of ocular abnormalities in our group of children. Our institution is a tertiary health care facility. Thus, referrals of the more severe and complicated cases are more common. This skewed population bias is to be seriously considered when evaluating the high rate of ocular malformations found in our IVF population. The presence of ROP in 2 of the children is easily reconcilable with the observations derived from the non-IVF population of premature infants.^{11 - 13} However, the many other serious structural malformations, which included chorioretinal and optic nerve colobomata, hypoplastic optic nerves, glaucoma, cataract, and retinoblastoma, deserve thorough consideration. The facts that most of these structural abnormalities were uniocular and there was no consanguinity or positive family history can be interpreted to indicate that the pathologic condition is sporadic and not due to an inherited trait. Taking into account the known higher rate of chromosomal aberrations in children born after IVF,^{8 - 9} the observations made in this study should be weighed seriously. A suggestion that these infants may experience a higher rate of ocular (and other organs) structural abnormalities is inevitable. Consequently, although the prevalence of 26% of ocular pathologic conditions in our cohort most probably does not represent the true expected incidence of ocular abnormalities in IVF offspring in general or their incidence in the population, it suggests that a higher incidence of structural abnormalities is to be expected. Minor structural abnormalities in other organs may remain "silent" and undetected. In the eye, however, not only are structural abnormalities readily detected, they also cause an impairment of visual function.

One of the children in our study developed a unilateral retinoblastoma. The possibility that this manifestation is due to a de novo deletion occurring as a "second hit" during early embryogenesis, as it occurs in the population of children at large, may be a plausible explanation for this phenomenon. An alternative reason for the occurrence of unilateral retinoblastoma may be that the 2-hit deletions have occurred in the retinal cells due to the higher tendency of chromosomal breakage and deletions in IVF-born children.^{8 - 9} Unfortunately, a closer examination of the problem was not possible in this case because the IVF was carried out with a donor sperm. This has precluded us from collecting useful genetic data, which could be of help for the elucidation of this enigmatic case.

The structural abnormalities observed in the children born after IVF have not been detected in utero. All mothers underwent periodic ultrasound surveys of the fetuses during the entire period of gestation. Modern ultrasound techniques and instruments can detect ocular abnormalities in utero.¹⁴ This is particularly applicable for microphthalmia,¹⁵ retinoblastoma,¹⁶ and congenital cataract.^{17 - 19} The lack of prenatal detection of these major ocular structural abnormalities in our study group may be attributed to the fact that the eyes are not carefully and routinely surveyed during ultrasound examination, leaving these malformations veiled.

Despite the limitations of our study, which include its retrospective design, lack of randomization, and incomplete obstetric survey, the results of our study should stimulate a heightened awareness about the possible higher incidence of ocular malformations in children born after IVF. Detection of these ocular structural malformations in high-risk pregnancies, such as after IVF in particular and in uncomplicated pregnancies in general, is feasible. Ophthalmologists and, particularly, obstetricians and gynecologists must be aware of the potential increased incidence of ocular abnormalities in IVF offspring. Early detection and treatment may be sight saving and are mandatory.

Accepted for publication May 24, 2001.

Corresponding author: Irene Anteby, MD, Pediatric Ophthalmology Unit, Department of Ophthalmology, Hadassah University Hospital, PO Box 12000, Jerusalem 91120, Israel (e-mail: irene@hadassah.org.il).