LABORATORY SCIENCES Evaluation of Normal Human Foveal Development Using Optical Coherence Tomography and Histologic Examination Adam M. Dubis, PhD; Deborah M. Costakos, MD, MSc; C. Devika Subramaniam, MD; Pooja Godara, MD; William J. Wirostko, MD; Joseph Carroll, PhD; Jan M. Provis, PhD Objective: To assess outer retinal layer maturation dur- bands was evident across the foveal region. These fea- ing late gestation and early postnatal life using optical tures are consistent with previous results from histologic coherence tomography and histologic examination. examinations. A “temporal divot” was present in some in- fants, and the foveal pit morphologic structure and the ex- Methods: Thirty-nine participants 30 weeks’ postmen- tent of inner retinal excavation were variable. strual age or older were imaged using a handheld opti- cal coherence tomography system, for a total of 102 Conclusions: Handheld optical coherence tomography imaging sessions. Foveal images from 16 participants (21 is a viable technique for evaluating neonatal retinas. In imaging sessions) were normal and evaluated for inner premature infants who do not develop retinopathy of pre- retinal excavation and the presence of outer retinal re- maturity, the foveal region seems to follow a develop- flective bands. Reflectivity profiles of central, parafo- mental time course similar to that associated with in utero veal, and parafoveal retina were extracted and were com- maturation. pared with age-matched histologic sections. Clinical Relevance: As pediatric optical coherence to- Results: The foveal pit morphologic structure in infants mography becomes more common, a better understand- was generally distinguishable from that in adults. Reflec- ing of normal foveal and macular development is needed. tivity profiles showed a single hyperreflective band at the Longitudinal imaging offers the opportunity to track post- fovea in all the infants younger than 42 weeks’ postmen- natal foveal development among preterm infants in whom strual age. Multiple bands were distinguishable in the outer poor visual outcomes are anticipated or to follow up treat- retina at the peri fovea by 32 weeks’ postmenstrual age and ment outcomes in this population. at the fovea by 3 months’ postterm. By 17 months’ post- natal, the characteristic appearance of 4 hyperreflective Arch Ophthalmol. 2012;130(10):1291-1300 HE FOVEA CENTRALIS IS A quent gestational changes include cen- feature of all anthropoid trifugal migration of inner retinal neu- primate retinas. The fo- rons, resulting in progressive reduction of Author Affiliations: veal avascular zone (FAZ), Author Affil Departments of Cell Biology, Departments increased cone density, and CME available online at Neurobiology, and Anatomy excavation of inner retinal neurons char- Neurobiolog (Drs Dubis and Carroll), T www.jamaarchivescme.com (Drs Dubis a acterize the foveal region. The most sa- Ophthalmology (Drs Costakos, and questions on page 1250 Ophthalmolo Subramaniam, Godara, lient feature of the human fovea is the shal- Subramaniam Wirostko, and Carroll), and low pit that is left behind by the lateral the ganglion and bipolar cell layers at the Wirostko, an Biophysics (Dr Carroll), displacement of the inner retinal layers (fo- Biophysics ( fovea. During the last trimester, photore- Medical College of Wisconsin, vea is Latin for pit). Before and during fo- Medical Coll ceptors are immature at the foveal center and Children’s Hospital of veal formation, cones become tightly and Children Wisconsin (Dr Costakos), compared with those at parafoveal and Wisconsin (D Milwaukee; and Australian packed, elongate, and migrate centrip- 13 Milwaukee; 1-3 parafoveal locations. During this pe- Research Council Centre of etally. The processes of cone packing and Research Co Excellence in Vision Science, pit development in humans have been ex- riod, the retinal pigment epithelium (RPE) Excellence in The John Curtin School of amined in histologic studies.1-7 Whereas also develops interdigitations with the im- The John Cu 1,2,7 Medical Research and the site of the future fovea can be identi- mature outer segments. Postnatally, the Medical Rese Australian National University fied as early as 12 weeks’ postmenstrual foveal pit contour continues to be modi- Australian N Medical School, Australian age (PMA) using morphologic criteria3,8 fied by cellular migration, reaching ma- Medical Scho National University, Canberra 7,9-12 2 National Un (Dr Provis). Dr Dubis is now and molecular cues, the pit itself does turity at roughly 18 months’ postterm. (Dr Provis). with Duke Eye Center, Durham, not emerge until after the FAZ is defined Photoreceptor elongation and displace- with Duke E North Carolina. at around 24 to 26 weeks’ PMA. Subse- ment into the fovea also continue postna- North Caroli ARCH OPHTHALMOL / VOL 130 (NO. 10), OCT 2012 WWW.ARCHOPHTHALMOL.COM 1291 ©2012 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/25/2021 tally, and the results of studies1,2 indicate that foveal cone HANDHELD PROBE SPECTRAL-DOMAIN OCT density reaches the lower range reported for adults by age 4 years. The exact relationship between the foveal Optical coherence tomography was performed using a hand- pit development and the photoreceptor mosaic matura- held probe spectral-domain OCT system (HHP-SDOCT; Biop- tion remains unclear.3,14,15 tigen). Reference arm position and instrument focus were ini- tially set on the basis of information provided by Maldonado Optical coherence tomography (OCT) enables visu- 24 alization of the retinal lamina and the organization of the et al, with additional adjustments performed if required to im- 16-18 prove image quality. The nominal image size for all the par- outer retina. Previously, the use of OCT had been lim- ticipants was 8ϫ8 mm, with an image density of 1000 A- ited to cooperative and fixating individuals. However, the scans per 100 B-scans (except for participant DC_0368, for recent availability of handheld systems enables imaging whom the nominal image size was a 6ϫ6-mm cube). The set- of pediatric populations, with clinical applications that ting assumes an axial length of 24 mm; however, given the include pediatric abusive head trauma19,20 and retinopa- shorter axial length of our population, the actual image length thy of prematurity (ROP).21-23 Recent work from one group is less (approximately 6 mm). Individual variation in axial length has aimed to increase the efficiency of pediatric OCT, pro- would further alter the image length for each participant. This viding some insights into normal development.24,25 Ad- provides a slightly different lateral resolution for each indi- vances in OCT image acquisition,26,27 with reinvestiga- vidual; however, this image size was selected to cover the maxi- tion by anatomists,17,28 have raised questions about the mum amount of retinal area (allowing visualization of the macula and the edge of the optic nerve in a single image). Images were initial anatomical assignment of layers observed with 16,18,25,26 exported in an audiovisual interleaving format using on- OCT. Defining the correlation between OCT bands instrument software (Bioptigen In Vivo Vue Clinic; Biopti- and photoreceptor structure is important for understand- gen), and frames containing photographs of the macula were ing foveal development. extracted. Using handheld OCT, we examined postnatal fea- tures of foveal development in a series of pediatric pa- HISTOLOGIC EXAMINATION tients.1,2,7,8,23,25 Across individuals of different ages, we compared emerging retinal layers in OCT with those in Histologic images were selected from a database assembled from age-matched histologic sections. We used these com- human and macaque retinal histologic sections collected by one parisons and information available in the literature17,18 of us (J.M.P.) and by Anita Hendrickson, PhD, at the Depart- to define reflectance bands in pediatric OCT images. We ment of Biological Structure, University of Washington, Se- also examined the variability in the foveal pit morpho- attle. Montages were assembled from high-resolution images ϫ logic structure among this pediatric population. A bet- obtained using a digital camera (HP; Carl Zeiss) through an 25 ter understanding of what is visualized by OCT will en- objective lens (Carl Zeiss), at a final optical magnification of 312.5ϫ, between January 1, 1995, and January 27, 2012, at the able reliable interpretation of in vivo developmental data University of Washington or at the Australian National Uni- and will facilitate the identification of normal vs patho- versity, Canberra. Sections were montages created by one of logic features in a clinical setting. us (J.M.P.) using available software (Photoshop CS5; Adobe Systems Incorporated). Macaque tissue was paraffin embed- METHODS ded, while all the human samples were embedded in methac- rylate; both sample types were stained with Richardson stain. PARTICIPANTS AGE-MATCHED AND LAYER-MATCHED Research on humans followed the tenets of the Declaration of HISTOLOGIC SECTIONS AND OCT IMAGES Helsinki and was approved by institutional review boards at the Medical College of Wisconsin and at Children’s Hospital of Wis- Optical coherence tomography images were age matched with consin. Informed consent was obtained from parents of all the 3 histologic images of human retinas at 35 weeks’ PMA, 41 weeks’ minors after explanation of the nature and possible conse- PMA, and 13 years’ postterm. Because of the rarity of human quences of the study. One adult participant provided his own donor retinas between 25 and 40 weeks’ PMA, we selected im- consent. Thirty-nine participants (26 male and 13 female) 30 ages from macaque retinas at fetal day 131 as approximately weeks’ PMA or older were recruited from local communities sur- equivalent to the human retina at 32 weeks’ PMA. To establish rounding the Medical College of Wisconsin. Thirty-five partici- a valid comparative time point, we used the cecal period (CP) pants were recruited from the neonatal intensive care unit at Chil- (the period between conception and eye opening) as a unit of dren’s Hospital of Wisconsin for an examination using anesthesia, development.