Retinoblastoma is the most common intraocular malignancy in childhood. At the turn of the century, nearly all cases were fatal. However, advances in diagnosis and treatment have resulted in a present survival rate of nearly 95%.1 Early detection of retinoblastoma combined with improved treatment modalities has been an important factor in the reduction of morbidity and mortality.
In this month's Pediatrics electronic pages, Canzano and Handa2 report on the results of pupillary dilation to improve the detection of leukocoria, the most common presenting sign in retinoblastoma. Seven consecutive children aged 2 days to 20 months with suspected retinoblastoma were examined by an ophthalmologist. The red reflex was evaluated for evidence of leukocoria before and after pupillary dilation. The authors report the detection of leukocoria in only 3 out of 10 undilated eyes with confirmed retinoblastoma. However, after pupillary dilation all the affected eyes were noted to have leukocoria. The authors recommend that pediatricians perform pupillary dilation whenever the possibility of retinoblastoma or leukocoria is considered. They also reference the use of pupillary dilation by pediatricians for routine screening of new patients.
The stated purpose of this study is to show that the detection of leukocoria with a direct ophthalmoscope is improved by dilating the pupil. It is well-known to ophthalmologists that pupillary dilation enhances visualization of the red reflex. The red reflex is often utilized in the evaluation of lens opacities and its usefulness is largely dependent on pupil size. Therefore, it is not surprising that the authors were able to improve their identification of an abnormal red reflex by dilating the pupil. Our concerns with this study are whether it provides enough evidence to support a role for pupillary dilation by the pediatrician and how it would be incorporated into pediatric practice.
Our first concern is with the design of this study and the potential for bias. The examiner is an ophthalmologist, well-experienced in identifying subtle changes in the red reflex. In addition, each patient already has a known suspicion for retinoblastoma. One of the patients had an eye previously enucleated because of retinoblastoma. These factors increase the likelihood and expectation of identifying an abnormal red reflex in the study population. The authors describe examination of the red reflex at multiple angles to assess the retinal quadrants for evidence of leukocoria. Would the same type of evaluation have been performed in the absence of clinical suspicion for retinoblastoma? The presence of retinoblastoma is quite rare in a pediatric practice. Because of the low incidence, many pediatricians may not encounter a child with retinoblastoma during their career. Thus, it is difficult to compare the detection of leukocoria by an ophthalmologist who has previous knowledge that 1 eye may have retinoblastoma with a pediatrician in general practice.
Our second concern is the examination technique used in this study. The undilated examinations were performed with room lights illuminated to maintain constriction of the pupils. This is unfortunate because the red reflex is normally evaluated in dim light to maximize pupil size. We wonder how many undilated eyes with indeterminate red reflexes or normal red reflexes would have been identified as having leukocoria by simply turning off the room lights. Dilating the pupil may not have been necessary to detect leukocoria.
Another concern regarding technique is the way the pupil was examined with the direct ophthalmoscope. It is described as holding the ophthalmoscope at arm's length. Normally, the instrument is positioned close to the examiner's eye and the pupil is visualized at arm's length.
We agree that pediatricians can develop excellent skills for the detection of leukocoria. We also agree that a constricted pupil can prevent visualization of the red reflex. What then, constitutes an appropriate use of pupillary dilation by the pediatrician? The authors recommend pupillary dilation when the possibility of retinoblastoma or leukocoria is considered. They also emphasize appropriately that a normal red reflex after pupillary dilation does not rule out retinoblastoma or other retinal disorders. The diagnosis of retinoblastoma requires examination of the retina using indirect ophthalmoscopy, often with scleral depression. Patients known to be at risk for retinoblastoma undergo repeated retinal examinations under anesthesia to detect new tumors as early as possible. Other causes of leukocoria also require complete ophthalmologic assessment for diagnosis. Thus, there does not appear to be any benefit in having pediatricians dilate patients with a suspicion of retinoblastoma or leukocoria. The fact that retinoblastoma can be present despite an apparently normal red reflex mandates referral to an ophthalmologist. Referral is indicated regardless of the pediatrician's assessment of the dilated red reflex. We would also question the utility of pediatricians examining the dilated eyes of uncooperative children with the direct ophthalmoscope. In these cases, the red reflex can remain extremely difficult to evaluate despite pupillary dilation and visualization of the retina with indirect ophthalmoscopy is necessary to ensure the absence of pathology.
The other indication for pupillary dilation referenced by the authors is the pediatrician's general examination. Several pediatric practices are cited where pupillary dilation is routinely performed within the first 6 weeks of life.3 It is important to consider the rationale and limitations of this type of screening. The goal of screening newborn infants is early detection of visual or life-threatening conditions. The only means available to pediatricians to assess the inside of the eye of a newborn infant is observation of the red reflex. The newborn infant is known to have relatively small pupils. Isenberg et al4 found an average pupil size of 3.6 × 0.9 mm measured in dim light in preterm and term infants. If the pupil size prevents evaluation of the red reflex, appropriate screening cannot be accomplished and pupillary dilation may be helpful. However, when the red reflex of the undilated eye is normal, the examiner can be confident that there are no significant congenital opacities of the cornea, lens or vitreous disrupting the visual axis. Thus, it is unnecessary to dilate the pupils of every newborn infant. Although a normal red reflex does not rule out retinoblastoma, does it justify dilating the pupils of every newborn infant?
A recent study of hereditary retinoblastoma attempted to determine the age of tumor development after a normal baseline examination at or near birth. Only 5% of cases were detected by 1 month of age. In contrast, 62% developed tumors by the age of 6 months.5 Although familial retinoblastoma is occasionally present at birth, most cases, hereditary or nonhereditary, develop months to years after birth.1 ,6 The incidence of retinoblastoma has been estimated to be between 1:17a̅000 and 1:34a̅000 live births; approximately 11 cases per 1 million in children under age 5.7 ,8 The low incidence combined with a more frequent toddler age of onset makes the likelihood of a pediatrician encountering retinoblastoma quite unlikely in the newborn infant. Pupillary dilation for the purpose of detecting retinoblastoma would be more appropriate at ages beyond the newborn period and at regular intervals. However, the low incidence of disease makes it impractical, especially because earlier detection of retinoblastoma has yet to be demonstrated.
Another concern with screening thousands and possibly millions of infants is the potential for side effects. Although the incidence of side effects associated with pupillary dilation is low, it is not zero. In addition to the adverse effects noted by the authors, including, fever, flushing, tachycardia, hypertension, and nausea and vomiting, other studies have reported seizures and psychotic reactions as well as intracranial hemorrhage in premature infants.9 ,10 If millions of infants undergo pupillary dilation, some of them will have serious side effects. If the incidence of side effects is 1 in 10 000 and 1 million infants are dilated, 100 infants will experience adverse reactions. That is close to half the number of estimated new cases of retinoblastoma each year.
Although leukocoria is the most common presenting sign of retinoblastoma, it is not always present. Developing tumors are initially transparent and do not produce leukocoria. Larger tumors can also be missed despite pupillary dilation. Tumors can grow very rapidly and may be found as early as 2 weeks after a normal examination.5 Only repeated retinal examinations with indirect ophthalmoscopy will reliably improve the detection of retinoblastoma.
The best opportunity to obtain earlier detection of retinoblastoma is by educating pediatricians and parents. Pediatricians should be reminded of the importance of examining the red reflex. Techniques should be reviewed, emphasizing reduced room illumination and the appropriate light intensity for the direct ophthalmoscope. The full strength light of the ophthalmoscope causes unwanted pupillary constriction and is not necessary to visualize the red reflex. Parental education should also be emphasized as part of the routine newborn examination. Written materials illustrating leukocoria and strabismus might be helpful. The importance of the parental history cannot be overemphasized. Parents have repeated eye contact with their children and can sometimes perceive subtle abnormalities in the pupillary light reflex or eye alignment. The findings may not be detectable in the office setting. Any questionable red reflex or ocular alignment abnormality observed by a parent deserves referral to an ophthalmologist.
Drs Canzano and Handa are to be congratulated for their efforts to improve the early detection of retinoblastoma. Pupillary dilation may be helpful to look for leukocoria and other red reflex changes when a small pupil prevents visualization of the red reflex. Additional studies are needed to show that routine pupillary dilation by pediatricians can enhance the detection of disease. Only then can it be recommended for routine screening.
- Received June 1, 1999.
- Accepted June 1, 1999.
Reprint requests to (L.B.N.) Wills Eye Hospital, Department of Pediatric Ophthalmology, 900 Walnut St, Philadelphia, PA 19107.
After July 2, 1999, Dr Metz's address will be: Department of Ophthalmology, Eye Foundation Hospital, University of Alabama/Birmingham, Birmingham, AL.
- ↵Canzano JC, Handa JT. Utility of pupillary dilation for detecting leukocoria in patients with retinoblastoma. Pediatrics. 1999;104;(4). URL: http://www.pediatrics.org/cgi/content/full/104/4/e44
- ↵Abramson DH, Frank CM, Susman M, Whalen MP, Dunkel IJ, Boyd NW III. Presenting signs of retinoblastoma. J Pediatr. 1998;132(3 pt 1): 505–508
- Copyright © 1999 American Academy of Pediatrics