When working with pediatric patients, it can be easy to fall into the trap of simply considering them small adults. From the ophthalmic examination to operating on complex pediatric pathology, some differences become apparent quickly, while some are subtle and require knowledge beforehand in order to tackle these unique cases. In this article, I’ll review some of the challenges of diagnosing and treating children with retinal conditions, and explain how to optimize outcomes as much as possible.

A Different Exam

The differential diagnosis for children with retinal pathology is different than it is for adults. The most common diagnoses we encounter in adult clinics are age-related macular degeneration, diabetic retinopathy, vein occlusions, macular hole, epiretinal membrane and retinal detachment. In pediatric retina, however, we take care of premature neonates with retinopathy of prematurity in the neonatal intensive care unit, older kids with traumatic retinal detachment, genetic conditions such as sickle cell retinopathy, Stickler syndrome, familial exudative vitreoretinopathy (FEVR), X-linked retino-schisis and incontinentia pigmenti, and sequelae of developmental anatomic anomalies like persistent fetal vasculature, optic disc pits and colobomas.

When diagnosing these conditions, in say, 3-year-olds running circles around the exam chair, the approach we take is quite different than a typical adult funduscopic exam. Neonates are typically easy to overpower, especially if they are swaddled. You will need additional hands to stabilize a child that’s approaching several months of age. Toys with bright lights and which make noises can also help children focus, and you can move these toys in the direction you would like them to look. 

For patients between 1 and 3, things get interesting, and your approach often depends on the child’s personality and developmental stage. First, you have to win the child’s trust, and you can’t go wrong with a high-five to initiate your interaction. I like to set the ophthalmoscope light low, and allow the child to sit on the parent’s lap. I quickly look at the red reflex first, as that alone can provide tons of information. If I suspect an area of pathology, I aim to look there first, since we don’t know when the child will decide that this isn’t fun anymore. Some children find it entertaining when I look for animals or cartoon characters inside their eyes. It usually helps to have the parents working with you also, as they may have the child’s favorite toys or tablet videos that they can use as distractions. When in doubt, and when faced with possibly blinding or life-threatening conditions, we must remember to have a low threshold to proceed with an examination under anesthesia.

Imaging technologies can also augment our ophthalmoscopic exam. This includes modalities such as widefield scanning laser ophthalmoscopy imaging and ultrasonography. SLO imaging, such as with the Optos camera, is quick, comfortable and great for imaging the peripheral retina, where many pediatric retinopathies may reside. A peripheral retinal exam is difficult even for adults to tolerate, and B-scan ultrasonography can supplement the peripheral ophthalmoscopic exam in a comfortable way in children. 

Figure 1. Widefield fluorescein angiography. (A) Widefield imaging captures the peripheral pathology in this child with familialexudative vitreoretinopathy. There is temporal dragging of the retinal vasculature and a peripheral traction retinal detachment without overlying hyaloidal organization. (B) Widefield fluorescein angiography demonstrates hyperfluorescence consistent with leakage from the vasculature. 

Figure 2: Oral fluorescein angiography. (A) For children who cannot tolerate intravenous injection of fluorescein, the dye can be
ingested to obtain late-phase angiograms. This is a child with sickle cell retinopathy. (B) Oral fluorescein angiography shows peripheral nonperfusion and neovascularization temporally.


Our understanding of the genetic basis of diseases continues to expand. Identifying the genetic cause for hereditary retinal diseases facilitates diagnosis, prognostication and how we follow and treat patients. However, genetic testing can be expensive, challenging in terms of insurance, and it can take weeks to months to obtain results. A cost-effective, quick and meaningful way to uncover a potential genetic condition is to examine family members. Not only will this approach provide a better understanding of your patient’s condition, but you may identify early pathology in family members that may be treatable.1

Another diagnostic gem for many pediatric retinal conditions is widefield fluorescein angiography (Figure 1).2,3 Many of the diagnoses listed at the top of the article are primary or secondary retinal vasculopathies, and detailed imaging of the peripheral retinal vasculature is key in making the correct diagnosis and subsequent treatment plan. It’s essential, for example, to identify and distinguish Coats’ disease vs. retinoblastoma vs. capillary hemangiomas.

However, many children won’t tolerate an intravenous injection of fluorescein. One useful trick, then, is to have children drink the dye, and then take the images approximately 20 minutes later. Oral FA is a great way to acquire late-phase angiograms (Figure 2).4 Note that you can’t obtain early and mid frames, so the diagnostic capacities are relatively limited compared to intravenous FA. For children too young for imaging in clinic, widefield FA can be performed under anesthesia. Also note that fluorescein is dosed by weight in children.5

Surgical Considerations

In pediatric retina surgery, remember two important concepts: 1) a young child’s ocular anatomy is different from an adult’s; and 2) the child’s vitreous is very formed and adherent to the retinal surface. Here’s how these factors can affect surgery, and how to account for them.

• Anatomic differences. The proportions of the ocular structures in young children are different from an adult’s—for instance, the lens is relatively large and the pars plana is small (Figure 3). The extent of the difference depends on the patient’s age. For instance, if we created a typical sclerotomy 3.5 to 4 mm posterior to the limbus in a neonate as we do for adults during vitrectomy, we would slice through retina and the result would be disastrous. Typically, for children younger than 1, we enter 1 mm posterior to the limbus; for kids between 1 and 2 years of age, we enter 1 to 2 mm posterior; and for children who are between 2 and 3 years old, we enter 2 to 3 mm posterior to the limbus. We enter 3 mm posterior for most other children. These are rough guidelines, and we always examine the eye carefully beforehand to make sure the pars plana is available for entry. This is because the pediatric patient’s ocular anatomy can be altered by the underlying pathology. A prime example is persistent fetal vasculature. There can be segments of retina that developed, or became pulled, very anteriorly by the ciliary processes. If we go through the pars plana in those areas, we could nick the retina and cause an inoperable rhegmatogenous retinal detachment.  

While pediatric eyes are smaller, their crystalline lenses are proportionately larger. We always have to be cognizant of the lens as we work in the tight intraocular space (unless we are removing the lens for very anterior retinal detachments). To decrease the risk of lenticular trauma, in the beginning of the surgery we aim our instruments more posteriorly, but away from any anteriorized retina. It’s also a good practice to slow down and be deliberate when inserting and removing instruments, especially curved ones.

Figure 3. Thoughtful trocar placement for young children. This 2-month-old with familial exudative vitreoretinopathy is undergoing vitrectomy. Note that the sclerotomies are made directly in the sclera without cannulas, because the surgical space is limited. In this patient, the sclerotomies are created 1 mm posterior to the limbus, unlike the usual 3.5 to 4 mm in older patients. The temporal sclerotomy is also displaced superiorly, to avoid an anterior tractional retinal detachment in the superotemporal quadrant.

• Visualization. One of the practical issues with young children is that their eyes are smaller, so we have to hold our instruments closer together. As we note above, the sclerotomies are created closer to the limbus, so that again decreases the distance between our right and left hands. Because we hold our instruments closer together, it’s easier to hit the viewing lens with our instruments. If we hit the viewing lens, we lose our view—and if we lose our view, we can’t operate. To help with this situation, there are certain non-contact viewing systems that allow the lens to sit higher off of the cornea. There are also specific pediatric retinal surgery lenses that are smaller in diameter than their adult counterparts.

• Scleral buckles are your best friends. For most vitreoretinal surgeons, vitrectomy is now the go-to method for repairing primary rhegmatogenous retinal detachment in adults. Vitrectomy usually works in adults because you can readily separate the vitreous from the retinal surface, followed by flattening of the retina, retinopexy and tamponade. In children, however, it can be impossible to completely remove the vitreous. An RRD repair without addressing the traction and without removing the scaffold for proliferative vitreoretinopathy means a lower success rate. Children are less likely to adhere to proper postop positioning as well, which means that you can’t count on the tamponade to work properly. 

It’s for these reasons that we prefer scleral buckling to fix RRDs in children (Figure 4). Scleral buckles don’t rely on the vitreous being separated, and you don’t have to rely on positioning. In fact, almost all primary RRDs in children would do great with a nicely placed scleral buckle. Primary buckling is a fading art, but remains a mainstay of pediatric RD surgery.5

• No margin for error. The child’s formed vitreous poses challenges for tractional retinal detachments also. In adults with diabetic TRDs or PVR retinal detachments, the general approach is to release the membranes, but if the membranes are too intrinsic to the retina and can’t be peeled, you can perform a retinectomy (especially for PVR) to cut the stiff retina and flatten the retina with brute force. You can’t take that approach in children, though. For a child with a TRD from conditions like ROP, Norrie disease or FEVR, the goal is to release all the traction without making any retinal breaks, and certainly no retinectomies. This is because in an adult you can almost always flatten the retina by resorting to cutting it and being aggressive with the peeling. In children, however, it’s almost impossible to flatten the retina completely due to the adherent vitreous and intense membranes, so a small iatrogenic retinal break in a TRD in a young child can lead to an inoperable eye. 

Since the stakes are high in these pediatric TRD surgeries, the old saying, “perfect is the enemy of good” couldn’t be any more true. The goal is to release the traction without making breaks, and then rely on the retinal pigment epithelium to gradually pump the subretinal fluid out and reattach the retina over time. If you must drain subretinal fluid or hemorrhage, it’s a good idea to drain externally through a sclerotomy, as opposed to the internal retinotomies that we’re used to making in adults. 

Figure 4. The formed and adherent vitreous, as well as the inability to position reliably, make scleral buckle the ideal surgical modality to treat rhegmatogenous retinal detachment in children. This child had an open globe repair and subsequently developed a retinal detachment, which was repaired with a primary encircling buckle.

• Think of the entire patient, not just the eyes. Some children and adults are at very high risk for anesthesia-related morbidity and mortality. Perhaps the highest risk as a general group would be ROP infants, who are likely to have multiple severe co-morbidities.2 These children also tend to have bilateral, progressive pathology. Over the course of several days, an infant with ROP can progress from 20/20 potential as a stage 4A retinal detachment to hand-motion vision as a stage 4B detachment. So how do we address bilateral, progressive disease in sick patients? Do we schedule two separate surgeries and potentially double the anesthesia risk, while watching the second eye progress? 

Bilateral ophthalmic surgery is usually not performed in the United States. The main theoretical risk is bilateral endophthalmitis, which indeed would be devastating. However, the risk for endophthalmitis after vitrectomy has been reported to be 0.03 percent to 0.08 percent, and assuming that we treat each eye as being independent of the other by redraping, rescrubbing, regowning, etc., the risk for bilateral endophthalmitis is equivalent to 1 patient in 1,500,000 to 10,000,000 bilateral surgeries.6 On the other hand, the risk for anesthesia-related mortality in children is as high as 1 in 10,000, 10 to 100 times higher for all-cause mortality; it’s even higher in neonates, and higher still in premature neonates.6 So the risk of mortality with a second anesthesia session is many times higher than the risk of developing bilateral endophthalmitis. 

In these instances, we recommend considering simultaneous bilateral surgery to decrease the risk of mortality and progression of blinding disease. We have previously published guidelines for this practice.6 We love focusing on patients’ eyes, but we also need to take a step back when our patients are ill and assess the entire patient to optimize outcomes—for both the vision and the life of the patient. 

Ultimately, we have to approach pediatric patients and their unique eyes thoughtfully. Pediatric retina surgery is a gratifying field, because you’re potentially saving vision for the next 100 years of the young patient’s life. There’s nothing like telling a mother that her baby is going to be able to see.  REVIEW


Dr. Yonekawa is a retina surgeon at Massachusetts Eye and Ear and directs the pediatric retina surgery program at Boston Children’s Hospital. He will be joining Mid Atlantic Retina and Wills Eye Hospital in Philadelphia later this year.


1. Kashani AH, Learned D, Nudleman E, et al. High prevalence of peripheral retinal vascular anomalies in family members of patients with FEVR. Ophthalmology 2014;121:1:262-268. 

2. Yonekawa Y, Thomas BJ, Thanos A, et al. The cutting edge of retinopathy of prematurity care: Expanding the boundaries of diagnosis and treatment. Retina 2017;37:12:2208-2225.

3. Kang KB, Wessel MM, Tong J, et al. Ultra-widefield imaging for the management of pediatric retinal diseases. J Pediatr Ophthalmol Strabismus 2013;50:5:282-288.

4. Nayak BK, Ghose S. A method for fundus evaluation in children with oral fluorescein. Br J Ophthalmol 1987;71:12:908-909.

5. Yonekawa Y, Fine HF. Practical pearls in pediatric vitreoretinal surgery. Ophthalmic Surg Lasers Imaging Retina 2018;49:8:561.

6. Yonekawa Y, Wu WC, Kusaka S, et al. Immediate sequential bilateral pediatric vitreoretinal surgery: An international multicenter study. Ophthalmology 2016;123:8:1802-1808.