Yale University
Yale University
EliScholar – A Digital Platform for Scholarly Publishing at Yale
EliScholar – A Digital Platform for Scholarly Publishing at Yale
Yale Medicine Thesis Digital Library
School of Medicine
1-1-2019
Multimodal Imaging And Asymmetry Of Disease Progression In
Multimodal Imaging And Asymmetry Of Disease Progression In
Rhodopsin-Associated Autosomal Dominant Retinitis Pigmentosa
Rhodopsin-Associated Autosomal Dominant Retinitis Pigmentosa
Lawrence Chan
Follow this and additional works at: https://elischolar.library.yale.edu/ymtdl
Part of the Medicine and Health Sciences Commons
Recommended Citation
Recommended Citation
Chan, Lawrence, “Multimodal Imaging And Asymmetry Of Disease Progression In Rhodopsin-Associated
Autosomal Dominant Retinitis Pigmentosa” (2019). Yale Medicine Thesis Digital Library. 3480.
https://elischolar.library.yale.edu/ymtdl/3480
This Open Access Thesis is brought to you for free and open access by the School of Medicine at EliScholar – A
Digital Platform for Scholarly Publishing at Yale. It has been accepted for inclusion in Yale Medicine Thesis Digital
Library by an authorized administrator of EliScholar – A Digital Platform for Scholarly Publishing at Yale. For more
information, please contact elischolar@yale.edu.
Multimodal Imaging and Asymmetry of Disease Progression in Rhodopsin-
associated Autosomal Dominant Retinitis Pigmentosa
A Thesis Submitted to the Yale University School of Medicine
in Partial Fulfillment of the Requirements for the Degree of Doctor of Medicine
by
Lawrence Chan
2019
Abstract
Retinitis pigmentosa (RP) is a group of genetically and clinically heterogeneous inherited
retinal degenerative diseases with no known cure to date. The recent gene therapy treatment
for Leber’s congenital amaurosis and RP caused by mutations in RPE65 have resulted in
dramatic improvements in vision, leading to excitement for other potential gene therapies on
the horizon. Upcoming clinical trials will be targeting patients with specific mutations, and
measurements of disease progression will be needed for each genetic subtype of RP in
order to determine whether treatments are successful. In this retrospective cohort study, we
examined 27 RP patients with confirmed autosomal dominant mutations in the rhodopsin
gene by monitoring rates of progression as measured structurally with ellipsoid zone (EZ)
line width on spectral domain optical coherence tomography (SD-OCT), horizontal and
vertical hyperautofluorescent ring diameters on short wavelength fundus autofluorescence
(SW-FAF), and as measured functionally with 30 Hz flicker amplitudes on
electroretinography (ERG). Each structural parameter was measured twice by the author
four weeks apart. The mean rates of progression were -158.5 μm per year (-8.4%) for EZ
line widths, -122.7 μm per year (-3.5%) for horizontal diameters, and -108.3 μm per year
(-3.9%) for vertical diameters. High test-retest reliability was observed for the parameters (EZ
line intraclass coefficient [ICC] = 0.9989, horizontal diameter ICC = 0.9889, vertical diameter
ICC = 0.9771). The three parameters were also correlated with each other (r = 0.9325 for EZ
line and horizontal diameter; r = 0.9081 for EZ line and vertical diameter; r = 0.9630 for
horizontal and vertical diameters). No significant changes in ERG amplitude were seen. The
subjects were classified by rhodopsin mutation class (I, IIa, IIb, III) and morphology of the
hyperautofluorescent ring (typical vs. atypical). No significant differences in rates of structural
progression were observed by rhodopsin mutation class or by ring morphology. Finally,
higher rates of asymmetry of progression between the left and right eyes were detected for
EZ line width (23% of subjects), horizontal diameter (17%), and vertical diameter (25%), as
compared to studies on other forms of RP.
Acknowledgments
I would like to thank my mentors and thesis advisors Dr. Stephen Tsang and Dr. Ron
Adelman for their mentorship and guidance with this project and my path through
medicine and ophthalmology. I would also like to thank members of the Tsang lab for
their invaluable assistance, including Dr. Ronaldo Carvalho for his help with planning the
experimental design and Jimmy Duong for his much-needed statistical wizardry and
patience with my incompetence. Furthermore, I also thank Dr. Ching-Hwa Sung from
Weill Cornell Medicine for her expertise in biochemical characterization of rhodopsin
mutations. I would like to express my gratitude to Dr. Ninani Kombo for her efforts in
helping me with the revision process. I also want to show my deepest appreciation to the
Yale Department of Ophthalmology and Visual Science, especially to Deana Ralston for
her incredible guidance with finalizing the thesis. I thank my wonderful friends,
classmates, mentors, and family for their unending support for as long as I can
remember. Finally, I want to thank my fiancée and life partner Yue Meng for her
unconditional love and guidance at every step of my life these past seven years.
Table of Contents
Introduction………………………………………………………………………………1
Statement of Purpose…………………………………………………………………16
Methods…………………………………………………………………………………17
Results……………………………………………………………………………………25
Discussion………………………………………………………………………………39
References……………………………………………………………………………..47
1
Introduction
Retinitis pigmentosa (RP), a group of inherited retinal diseases with an incidence
of approximately one in 4000 people, is characterized by progressive
photoreceptor death and irreversible vision loss (1). Typically, the initial loss of
photoreceptors primarily involves the rods, thereby diminishing peripheral and
night vision, followed by worsening tunnel vision and eventual loss of central
vision mediated by cone photoreceptor death (1). Ophthalmoscopic hallmarks of
the disease include retinal arteriolar attenuation, bone-spicule peripheral pigment
deposits, and waxy pallor of the optic disc (2). The clinical presentation of retinitis
pigmentosa is highly variable. The severity and pattern of vision loss may be mild
or severe. The rate of disease progression can be slow or rapid, and the age of
onset can be as early as childhood while some individuals remain asymptomatic
until mid-adulthood. Allelic heterogeneity, in which each gene locus may have
different mutations that cause the same disease entity, contributes to the diverse
genetic etiology of RP; for example, over 300 different RPGR mutations have
been identified in families with X-linked RP (3). Even among members of the
same family, the same mutation may result in different phenotypic
manifestations. RP is also a genetically heterogeneous disease, with over 50
genes that have been found to be associated with non-syndromic RP. Further
complicating the heterogeneity of the disease is that different mutations in the
same gene may result in different modes of inheritance. The pattern of
inheritance can be autosomal recessive (15-20%), autosomal dominant (20-
2
25%), X-linked recessive (10-15%), or sporadic (30%) (2, 4). RP may also be
syndromic, as seen in Bardet-Biedl syndrome, Usher syndrome,
abetalipoproteinemia (Bassen-Kornzweig syndrome), and phytanic acid oxidase
deficiency (Refsum disease) (2).
Despite the genetic complexity of RP, improvements in the cost and efficiency of
molecular techniques that allow for the high-throughput DNA sequencing of
patients have resulted in clinicians being able to append a molecular diagnosis to
their clinical diagnosis. Specifically, the advent of next-generation sequencing
(NGS), which is able to perform massively parallel sequencing runs on the order
of millions of DNA fragments using micron-sized beads, has dramatically
increased the speed of sequencing many-fold and enabled the capture of a
broader spectrum of mutations compared to conventional Sanger sequencing (5).
Molecular basis of the visual cycle
To understand how mutations in certain genes may cause RP, an outline of the
visual cycle will need to be described. The first step in vision occurs when light
enters the eye and is focused by the cornea and lens onto the retina
(photosensitive tissue located posteriorly within the eye). In the retina, the light-
sensitive photoreceptor cells called rods and cones convert the external light
stimuli into electrical impulses that the brain processes to form an image. Rod
photoreceptors contain the visual pigment rhodopsin, which is a light-sensitive G-
3
protein coupled receptor that consists of the apoprotein opsin and 11-cis-retinal,
a chromophore. When light is absorbed by rhodopsin, the 11-cis-retinal is
converted to all-trans-retinal and leads to a series of conformational changes of
the opsin that activates the GTP-binding protein transducin, triggering a
canonical cyclic guanosine monophosphate (cGMP) second-messenger cascade
through the activation of cGMP phosphodiesterase (PDE) (2). PDE hydrolyzes
cGMP, leading to closure of the cGMP-dependent cation channels normally
responsible for influx of Na+, Ca2+, and Mg2+. The resulting hyperpolarization of
the photoreceptor cell decreases the rate of transmitter release and elicits
responses in second-order (bipolar) cells for further neural transmission (6). The
all-trans-retinal is converted to all-trans-retinol and is transported to the retinal
pigment epithelium (RPE) to be recycled into 11-cis-retinal for transport back into
the rods (2).
Rods are sensitive to low levels of light, and psychophysical experiments have
shown that they can register single photon absorptions (6). Since rods play a
crucial role in enabling vision in low-light scenarios and are anatomically located
in the periphery of the retina, RP patients usually experience night blindness
(nyctalopia) and loss of peripheral vision as their initial symptoms.
The organization of the rod photoreceptor consists of a synaptic body that
interfaces with the bipolar/horizontal cells, a cell body, an inner segment (IS)
which contains the endoplasmic reticulum, mitochondria, and Golgi apparatus,
4
and an outer segment (OS) which houses membranous discs containing mostly
opsin within a plasma membrane. The IS and OS are connected by the
connecting cilium, and the OS interfaces with and is phagocytosed by the RPE.
Figure 1. a) Illustration showing cell organization within the retina. b) Cross-
sectional H&E stain of retina. Image from Wikimedia Commons.
5
Structure of rhodopsin
As previously mentioned, rhodopsin (RHO) is the G-protein coupled receptor
(GPCR) that is responsible for the first step in allowing rod photoreceptors to
detect light. It is synthesized in the rough endoplasmic reticulum and then
transported through the Golgi apparatus where it ultimately functions within the
discs of the OS (7). 30% to 40% of all autosomal dominant RP (adRP) is caused
by mutations in the RHO gene, and over 120 different mutations in RHO have
been identified (2, 8). One study of 200 families with clinical evidence of adRP
found that rhodopsin mutations were the most common cause of disease,
representing 26.5% of the total cases of adRP (9). In addition to its role in adRP,
rhodopsin was the first GPCR whose crystal structure was elucidated, and it
served as a prototype template for understanding the rest of the GPCR
superfamily (8). Rhodopsin is a highly conserved protein among vertebrate
species, and similar proteins have even been found in the visual systems of
invertebrates such as Drosophila melanogaster (10). The structure of rhodopsin
consists of four specialized domains that assist in the maintenance of protein
structure, trafficking, and phototransduction: 1) cytoplasmic, 2) intradiscal, 3)
transmembrane, and 4) ligand-binding domains (11). The cytoplasmic C-terminal
domain of rhodopsin regulates its trafficking and interactions with other proteins
in the phototransduction cascade such as transducin (11). The intradiscal domain
contains the extracellular loops between transmembrane domains and the N-
terminus. Research suggests that mutations in the intradiscal domain result in
6
misfolding of the protein and accumulation of the protein within the secretory
system, leading to disease (12). The transmembrane domains have been shown
to have several residues that are important for rhodopsin protein stability and
function (13). The ligand-binding domain is where the 11-cis-retinal chromophore
binds with the opsin apoprotein (14).
Biochemical classification of rhodopsin mutations
Mutations in rhodopsin causing adRP have been grouped into three classes
(Table 1) based on the phenotypes of the proteins from in vitro studies that
transfected human tissue culture cells with wild-type and mutant rhodopsin cDNA
clones (8, 11, 12). Class I mutations are located near the C-terminus of the
protein or within the first transmembrane segment. The protein resembles wild-
type rhodopsin in terms of protein levels, ability to associate with the 11-cis-
retinal chromophore, and subcellular localization (15, 16). However, these
mutations cause rhodopsin to activate transducin inefficiently in the presence of
light (17). Class II mutations cause decreased binding to 11-cis-retinal and result
in accumulation within the endoplasmic reticulum, possibly due to issues with
protein folding and stability (15, 17). Within class II, further subclassification can
be made for those mutants that predominantly localize intracellularly (class IIa)
and those that preferentially localize to the cell surface (class IIb) (16). Finally,
class III mutants form rhodopsin poorly and at low levels, are retained in the
endoplasmic reticulum, and may form aggresomes, causing targeted degradation
7
by the ubiquitin proteasome system (18). Studies have suggested that impaired
endocytic activity is the primary mechanism by which class III mutations cause
RP (19). One common finding among all three classes of mutations is the
decreased sensitivity to light and less efficient activation of transducin (17).
Table 1. Classification and description of rhodopsin mutants
Class
Biochemical phenotype
I
Mutations occur near C-terminus
Similar to wild-type rhodopsin
Inefficient activation of transducin
II
Misfolding/instability
Accumulation within endoplasmic reticulum
Class IIa: localize intracellularly
Class IIb: localize to cell surface
III
Impaired endocytosis from membrane
Form rhodopsin chromophore poorly
Accumulation within endoplasmic reticulum
Clinical classification of rhodopsin patients
Aside from the preceding classification of rhodopsin mutations based on
biochemical characteristics, research on adRP caused by rhodopsin mutations
has produced evidence of two different subtypes predicated on the clinical
pattern of disease. The class A phenotype, sometimes referred to as “type 1” or
“diffuse” subtypes, is characterized by a severe, early-onset diffuse loss of rod
sensitivity with a later prolonged degeneration of cones (20, 21). The class B
phenotype, also known as “type 2” or “regional”, exhibits a combined loss of rod
and cone sensitivity in a superior hemifield (altitudinal) pattern with relatively
preserved function in the inferior hemifield, as well as a slower progression of
8
disease with night blindness manifesting during adulthood (21). Because of the
regionalized retinal degeneration in the altitudinal pattern, these phenotypic
variants of RP are also known as sector RP (22). It has been postulated that the
more severe class A phenotype may be caused by a gain-of-function mutation
that is cytotoxic, while the milder class B phenotype is a result of a loss-of-
function mutation inherited on a single allele (23).
Potential for therapeutic intervention
Gene therapy is an experimental technique that seeks to treat genetic disorders
by replacing or supplementing the mutated gene with a healthy copy of the gene,
or inactivating a mutated gene, in contrast to traditional therapies such as
surgery and medications. In late 2017, Spark Therapeutics’ LUXTURNA™
(voretigene neparvovec), a treatment for LCA and RP caused by mutations in the
RPE65 gene, became the first gene therapy for any disease to gain regulatory
approval in the United States by the Food and Drug Administration. This gene
therapy involves the subretinal injection of wild-type copies of RPE65 packaged
in an adeno-associated virus (AAV).
RPE65 (retinal pigment epithelium-specific protein, 65 kDa) is responsible for
producing the isomerase enzyme that catalyzes the isomerization of all-trans-
retinal back to 11-cis-retinal within the retinal pigment epithelium so that the
previously mentioned visual cycle can begin again (24). In LCA and RP caused
9
by bi-allelic mutations in RPE65, the visual cycle is disrupted and photoreceptors
undergo dysfunction and degeneration, the two pathological mechanisms that
ultimately lead to progressive blindness (25). Early preclinical studies in mouse
and dog models have shown that gene augmentation therapy is able to correct
the biochemical blockade and result in significant, persistent vision improvement
(25). These promising initial results over the past two decades led to the
University of Pennsylvania research group to collaborate with Spark
Therapeutics to test the efficacy and safety of AAV2-hRPE65v2 (voretigene
neparvovec) on 31 patients across two leading US academic centers for the
study of inherited retinal dystrophies (Children’s Hospital of Philadelphia,
Philadelphia, PA and University of Iowa, Iowa City, IA). This randomized
controlled study, the first phase 3 trial for any gene therapy, demonstrated
clinically and statistically significant improvements in the subjects’ visual field
measurements and ability to independently navigate in low-light conditions,
persisting throughout the one-year follow-up period (26).
The success of the RPE65 gene therapy trials has spawned a large number of
clinical trials seeking to use gene therapy to cure other inherited retinal
degenerative diseases. For example, there are several endeavors in the US and
the UK to study gene therapy treatments for choroideremia, an X-linked
recessive retinal disease that causes progressive loss of peripheral vision and
night blindness (27, 28). Choroideremia is caused by mutations in the CHM
gene, which encodes for the Rab escort protein-1 (REP1). This condition is
10
amenable to treatment with gene therapy using an adeno-associated virus 2
(AAV2) capsid due to the relatively small size of the CHM cDNA payload that can
be contained with the AAV2 vector (28).
However, unlike the loss-of-function mutations of the recessive choroideremia
and LCA that can be addressed with simple replacement of the wild-type gene,
RP caused by a dominant RHO mutation acquires an abnormal gain of function
that requires suppression of the mutant RHO gene and replacement with the
wild-type version. Strategies for suppressing the toxic gene include
transcriptional silencing, RNA interference, and ablation or correction of the
mutation at the DNA level using gene editing techniques such as zinc finger
nucleases (ZFNs), transcription activator-like effector-based nucleases
(TALENs), and the recently discovered clustered regularly interspaced
palindromic repeats (CRISPR)/Cas9 system. For RHO-adRP specifically, efforts
over the past few decades have focused on either targeting specific mutant
alleles for reduction in expression levels or by implementing a mutation-
independent knockdown strategy (29-31). The mutation-independent strategy is
particularly useful given the heterogeneity of the disease due to the large number
of disease-causing RHO mutations. This generally involves silencing the
expression of both the mutant and wild-type RHO alleles, while supplementing
wild-type protein-encoding RHO cDNA that is modified to be resistant to the
suppressor. Various methods exist to silence gene expression, including RNA
interference (RNAi) via short hairpin RNA (shRNA) or small interfering RNA
11
(siRNA), CRISPR/Cas9, and TALENs (32). One way of conferring resistance to
the replacement RHO cDNA is to modify codons to contain wobble nucleotides at
the target site, thereby decreasing hybridization with the suppressor reagent (33).
The goal of finding treatments that are targeted to each specific genetic disorder
is timely given the launch of the United States Precision Medicine Initiative during
President Barack Obama’s tenure. There has been an increased interest among
the scientific and medical communities to discover “precision medicine”
treatments tailored to each individual’s variability in genes, environment, and
lifestyle (34). Given the inevitable progress within the next decade in the field of
gene therapy in the wake of LUXTURNA, there is a crucial obligation to
characterize the natural history progression of each disease on a gene-by-gene
basis. Without baseline measurements of disease progression rates and
asymmetry between eyes, it will be difficult to determine the efficacy of retinal
gene therapy even with an untreated control eye.
Structural and functional assessments
Various structural and functional measures of disease severity exist within the
field of ophthalmology. Visual acuity and visual field testing are able to capture
the patient’s perception of visual impairment, but they are subjective tests that
have low test-retest reliability (35, 36). An objective method of assessing visual
function is electroretinography (ERG). This noninvasive electrophysiologic test of
12
retinal function uses recording electrodes placed on the corneal surface and
measures the changes in electric potential (against a reference electrode placed
on the skin) in response to light stimuli of varying intensities under dark- and
light-adapted conditions. The stimulation of the retina produces characteristic
waveforms that provide information about the function of different cells within the
retina, such as rods, cones, bipolar cells, retinal ganglion cells, and amacrine
cells. Important parameters of the waveform include the a- and b-wave
amplitudes (distance from baseline to a-wave trough, and from a-wave trough to
b-wave peak, respectively) and implicit time (time between stimulus onset and
maximum amplitude). The ERG is a useful tool in diagnosing many retinal
conditions, including retinitis pigmentosa, congenital stationary night blindness,
achromatopsia, toxic retinopathies, and cancer-associated retinopathy (37). It
also has utility in objectively assessing the retinal function in animal research
models. There are different forms of ERG, such as the standardized full-field
ERG (ffERG) which measures the total retinal response, pattern ERG (PERG)
which assesses central retinal function, and the multifocal ERG (mfERG) which
can detect localized responses in precise regions of the retina within the central
30 degrees (38). ERGs have shown increased reproducibility of measurements
compared to visual field testing, but are limited in their ability to reliably detect
small variations such as in end-stage retinal disease (35).
Imaging modalities such as spectral-domain optical coherence tomography (SD-
OCT) and fundus autofluorescence (FAF) have also been shown to be practical
13
tools in providing data about retinal and RPE structures that correlate well with
disease progression and functional measures (39). With the loss of
photoreceptors in the periphery that gradually progresses towards the fovea seen
in RP, it is important to be able to visualize and differentiate between the
dysfunctional, diseased portions and the healthy viable regions of the retina. One
visual marker of this border is the parafoveal ring of increased autofluorescence
first shown to be correlated with PERG by Robson et al. in 2003 (40). The short-
wavelength autofluorescence (SW-AF) imaging technique uses blue light
excitation at 488 nm and detects signals originating from lipofuscin granules and
other fluorophores within the RPE/photoreceptor complex (41). In RP patients,
these signals may manifest as rings and are thought to be the transition between
healthy and diseased retinal areas, with normal function within the ring and
dysfunction outside the ring (42). Some researchers have theorized that the
increased intensity of the autofluorescence signal is due to atrophy or stress-
induced accumulation of lipofuscin – the oxidative byproduct of phagocytosed
photoreceptor outer segments – within the RPE (43). The maximum intensity of
the signal captured by FAF may therefore represent the distribution of active
degeneration of photoreceptors where there is a high rate of phagocytosis by the
RPE; dark areas seen on fundus autofluorescence are indicative of atrophy of
the RPE and corresponding loss of lipofuscin granules (44). Studies have
demonstrated that the rate of hyperautofluorescent ring constriction is correlated
with visual field loss progression and has prognostic value in predicting visual
field acuity and visual field preservation (45).
14
In addition to FAF, SD-OCT is another noninvasive imaging modality that can
allow for in vivo visualization of the retinal layers. One hyperreflective band layer
that can provide information about photoreceptor health and function is the
ellipsoid zone (EZ), previously known as the inner segment/outer segment
(IS/OS) line (though the precise anatomic origins continue to be a topic of
debate). The hyperreflectivity of the EZ likely corresponds with the light scattering
by the mitochondria within the distal portion of the inner segment (46). Disruption
and/or shortening of the EZ line width corresponds with loss of visual field
sensitivity and thus provides a structural marker for the visual field edge (47, 48).
Some studies have shown that measurement of the EZ line width may be more
sensitive than full-field ERG and standard visual field testing in detecting
progression of visual field changes in RP. Birch et al. found that the rate of
change in EZ line width is consistent with those reported for ERGs and visual
fields, yet the test-retest variability of the EZ line width was considerably lower
(39). Furthermore, Birch et al. showed that the edge of the EZ line is where the
visual field sensitivity changes most accurately, and that observing this region is
more sensitive in detecting disease progression than global measurements that
average across the entire field (i.e., monitoring the healthy macula and the
diseased periphery, which are relatively stable) (49).
15
Figure 2. SD-OCT image showing retinal layers. Red arrow heads pointing to ellipsoid zone line
layer (top). Measurement of ellipsoid zone line width between dotted lines (bottom).
16
Statement of Purpose
Retinitis pigmentosa (RP) is a group of inherited retinal degenerative diseases affecting
roughly one in 4000 people worldwide and manifests as a progressive loss of vision. The
pattern of visual loss generally involves the initial degeneration of the rod
photoreceptors, followed by loss of the cones. It is marked by clinical and genetic
heterogeneity, with varying rates of vision loss and levels of disease severity, different
modes of inheritance, and more than 100 genes whose mutations have been found to
cause RP. There is currently no known cure for RP, but the recent groundbreaking FDA-
approved gene therapy treatment (LUXTURNA™) for Leber’s congenital amaurosis and
RP caused by mutations in RPE65 has shown dramatic improvements in vision and
given promise that gene therapy is a viable strategy for treating inherited retinal
diseases.
For future gene therapy clinical trials, it will be crucial to have data regarding RP natural
disease history and appropriate outcome measurements on a gene-by-gene basis given
the heterogeneity of the disease. Furthermore, precise details about disease severity
based on the various types of mutations within a single gene would inform researchers
about their decisions to enroll patients with certain mutations. In this study, we seek to
examine a subset of autosomal dominant RP patients with known mutations in the
rhodopsin gene (RHO) using structural (ellipsoid zone line width, hyperautofluorescent
ring diameters) and functional (electroretinography) assessments to monitor disease
progression. We will also look for asymmetry of rates between eyes and any correlations
between the rhodopsin mutation class, morphology of the hyperautofluorescent ring, and
disease severity.
17
Methods
Subjects
This study was conducted in adherence to the tenets of the Declaration of
Helsinki. All study procedures were defined and approved by the Institutional
Review Board at the Edward Harkness Eye Institute and Columbia University
Medical Center (Protocol #AAAR0284). Patient consent was obtained from all
subjects. The patient data presented here, including images and genetic testing
results, are not identifiable to individual patients. Diagnoses of RP were made by
an inherited retinal disease specialist (S.H.T.) based on clinical history, fundus
examinations, and full-field electroretinography (ffERG) results. This is a
retrospective cohort study with the following inclusion criteria: 1) patients must
have genetic sequencing-confirmed RHO mutations; and 2) a complete
ophthalmic examination must have been performed by our retinal disease
specialist on at least one visit. Since our clinic is an international referral center
for patients with RP, a significant portion of the subjects had their care
transferred back to their primary provider after the initial diagnosis was made in
our clinic using imaging, electroretinography, and genetic testing and thus did not
return for a follow-up visit. Patients were excluded if they: 1) presented with
advanced stage RP with no visible ellipsoid zone line in any eye at all time
points; 2) had unilateral RP; 3) did not have any visible hyperautofluorescent ring
in any eye at all time points; and 4) had poor image quality. A total of 38 patients
fit our inclusion criteria; 11 patients were excluded based on the exclusion
18
criteria, leaving a total of 27 patients on whom to base our analysis. The 38
patients belonged to 21 different families; the final 27 subjects belonged to 18
different families. For the 27 patients who were studied, eyes were analyzed only
if there were visible EZ lines/hyperautofluorescent rings; if there were no EZ
lines/hyperautofluorescent rings, the eye at that time point was not included.
Genetic analysis
DNA was extracted from the blood obtained from patients and was tested for
previously published RP genes of the Chiang panel at Columbia University
Medical Center Department of Pathology and Oregon Health Sciences
University. Parallel sequencing was performed using the Illumina HiSeq platform
with 100 bp paired-end reads, and mutations were confirmed by dideoxy chain-
terminating sequencing.
Mutation classification
Each patient was assigned a biochemical rhodopsin mutation classification
based on PubMed literature searches for each specific mutation. Biochemical
classifications were found for 32 out of 38 patients. Patients 3 and 4 had
mutations that have not been studied and classified. For patients 6-9, the
mutations were studied in bovine rhodopsin and were not characterized using the
19
classification system proposed by Sung et al. (16, 50, 51). Table 2 lists the
mutation classifications as well as their corresponding literature references.
Table 2: Rhodopsin biochemical mutation classification, including excluded patients
Genotype
Mutation class
Reference
1
RHO (c.556T>C:p.Ser186Pro)
IIa
PMID8253795
2
RHO (c.937-27_-19delCCCTGACTC)
I
PMC52606
3
RHO (c.946delT:p.Cys316Alafs*44)
–
4
RHO (c.946delT:p.Cys316Alafs*44)
–
5
RHO (c.266G>A:p.Gly89Asp)
IIb
PMC52606
6
RHO (c.83A>G:p.Glu28Arg)
–
7
RHO (c.328T>C:p.Cys110Arg)
–
8
RHO (c.328T>C:p.Cys110Arg)
–
9
RHO (c.328T>C:p.Cys110Arg)
–
10
RHO (c.568G>A:p.Asp190Asn)
IIa
PMID8253795
11
RHO (c.568G>A:p.Asp190Asn)
IIa
PMID8253795
12
RHO (c.266G>Ap.Gly89Asp)
IIb
PMC52606
13
RHO (c.1025G>A:p.Thr342Met)
I
PMC52606
14
RHO (c.541G>A:p.Glu181Lys)
IIa
PMID8253795
15
RHO (c.800C>T:p.267Leu)
IIa
PMID8253795
16
RHO (c.316G>A:p.Gly106Arg)
IIb
PMID8253795
17
RHO (c.404G>T:p.Arg135Leu)
III
PMC437971
18
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
19
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
20
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
21
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
22
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
23
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
24
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
25
RHO (c.800C>T:p.267Leu)
IIa
PMID8253795
26
RHO (c.632A>C:p.His211Pro)
IIa
PMID8253795
27
RHO (c.50C>T:p.Thr17Met)
IIa
PMC52606
28
RHO (c.50C>T:p.Thr17Met)
IIa
PMC52606
29
RHO
(c.404_405delinsGG>TT:p.Arg135Leu)
III
PMC437971
30
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
31
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
32
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
33
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
34
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
35
RHO (c.68C>A:p.Pro23His)
IIa
PMC52606
36
RHO (c.403C>T:p.Arg135Trp)
III
PMC437971
37
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
38
RHO (c.1040C>T:p.Pro347Leu)
I
PMC52606
RHO = rhodopsin; – = mutation class unknown
20
Image acquisition and measurements
Imaging was conducted after adequate pupil dilation (>7 mm) using
phenylephrine hydrochloride (2.5%) and tropicamide (1%). Fundus
autofluorescence (FAF, 488 nm excitation) and horizontal 9 mm SD-OCT images
at the fovea were acquired using the Spectralis HRA+OCT (Heidelberg
Engineering, Heidelberg, Germany) at each visit. OCT imaging was assisted by
eye-tracking technology that enables accurate and reproducible scans at the
same location on the fovea across multiple visits. The images were recorded with
a 30-degree field of view; in cases where the rings were too large to be
visualized with the 30-degree field of view, scans with a 55-degree field of view
were also captured.
The ellipsoid zone line widths, and horizontal and vertical diameters of the
hyperautofluorescent ring were manually measured using the built-in measuring
tool provided by the Spectralis software. The ellipsoid zone line width was
measured between the nasal and temporal limits of the ellipsoid zone layer using
the horizontal foveal scan on SD-OCT. The external border of the
hyperautofluorescent ring was used to determine diameter length, as it is more
clearly defined and easily visualized compared to the internal border. The
horizontal diameter is oriented along the axis formed by the center of the fovea
and the center of the optic disc. The vertical diameter is defined as the length of
the line between the external ring border, perpendicular to the horizontal
