Rates of Visual Field Change in Eyes with Optic Disc Drusen

Optic disc drusen (ODD) are acellular deposits of calcium, mucopolysaccharides, and nucleic and amino acids in the prelaminar portion of the optic nerve head.¹ It has been postulated that these calcified bodies may cause mechanical compression of the retinal nerve fibers and decreased blood perfusion of the optic nerve head, placing eyes with ODD at risk for axonal loss and subsequent visual field defects.² Therefore, estimation of rates of functional change and the evaluation of factors associated with visual field progression are essential to guide the management and to give a prognosis to these patients.

To date, only one longitudinal study has evaluated rates of visual field loss in eyes with ODD. Lee et al³ reported a rate of loss of 1.6%/year in the Goldman visual field in a cohort of 60 eyes. However, while Goldman perimetry may perform well for qualitative evaluations of visual field defects, accurate and precise quantification of rates of visual field change with this technique may be very challenging. In fact, Goldman perimetry has been largely replaced by standard automated perimetry (SAP), which is more reproducible and now routinely used to follow patients with ODD and other optic neuropathies.

There is also a paucity of studies evaluating risk factors for visual field loss in eyes with ODD. Some authors have previously reported associations between intraocular pressure (IOP) and visual field defects in eyes with ODD and proposed that lowering IOP would prevent visual field loss.^4–6 However, most of these studies have been cross-sectional, with small samples and general lack of adjustment for potentially confounding factors.

In the present work, we evaluated rates of SAP visual field loss in a cohort of eyes with ODD and investigated their relationship with demographic, clinical factors, and levels of IOP during follow-up.

METHODS

This was a retrospective cohort study of patients from the Duke Ophthalmic Registry (DOR), a database of Electronic Health Records (EHR) developed by the Vision, Imaging and Performance (VIP) Laboratory.⁷ The database consists of adults 18 years or older who were evaluated at the Duke Eye Center or its satellite clinics between January 2009 and September 2019. The Duke University Institutional Review Board approved this study with a waiver of informed consent due to the retrospective nature of this work. All methods adhered to the tenets of the Declaration of Helsinki for research involving human subjects and were conducted in accordance with regulations of the Health Insurance Portability and Accountability Act.

The database used for this study contained clinical information from baseline and follow-up visits, including patient diagnostic and procedure codes, medical history, best-corrected visual acuity, slit-lamp biomicroscopy, IOP obtained using Tono-Pen or Goldman applanation tonometry (GAT), gonioscopy, ophthalmoscopy examination, stereoscopic optic disc photographs, and SAP acquired with the Humphrey Field Analyzer (HFA, versions II and III; Carl Zeiss Meditec, Inc., Dublin, CA).

All eyes with an International Classification of Diseases (ICD) code for ODD in the database were included (i.e., “377.21”, “H47.321”, H47.322, “H47.323”, and “H47.329”). A diagnosis of ODD was further confirmed by review of medical records based on reports of ophthalmoscopy, B-scan ultrasound, autofluorescence on preinjection fluorescein angiography, computerized tomography scan, and/or optical coherence tomography (OCT).

Eligible subjects were required to have at least 3 reliable SAP tests defined as any 30-2 Swedish Interactive Threshold Algorithm (SITA) tests or 24-2 SITA Standard or SITA Fast strategy tests with size III white stimulus, with less than 33% fixation losses and less than 15% of false positives, and a minimum of 12 months of follow-up. Mean IOP was calculated as the average of all IOP measurements obtained during follow-up, while peak IOP was the highest IOP value during follow-up. The study baseline visit was defined as the date of the first reliable visual field test.

Subjects were excluded if they had any concurrent ocular or systemic disease that could affect the afferent visual function (e.g., ischemic optic neuropathy, retinitis pigmentosa). Given the difficulty in differentiating visual field loss from optic nerve damage in ODD from that caused by glaucoma, we did not a priori exclude eyes that were considered glaucoma suspects or had a diagnosis of glaucoma according to clinicians’ notes. As we were interested in evaluating IOP effects of rates of visual field loss in eyes with ODD, we also did not use IOP as an exclusion criterion.

Statistical Analysis

Linear mixed models (LMMs) were used to estimate rates of SAP mean deviation (MD) loss over time. This standard technique has been described in detail elsewhere.⁸ In brief, mixed models take into account the natural correlation of such data over time, as well as the fact that each patient may contribute with 2 eyes for the analyses. Differences in rates of change between eyes and subjects are considered by introducing random slopes and random intercepts. Best linear unbiased prediction (BLUP) was used to estimate individual slopes of change for each eye. These estimates are more precise than those obtained by ordinary least squares linear regression notably in the presence of a small number of tests over time as can occur with some eyes. As the number of tests increases, BLUP estimates become essentially identical to those obtained from ordinary least squares regression.^{9, 10} Individual slopes of SAP MD change were classified according to pre-established cutoffs of slow, if the change was slower than −0.5 dB/year; moderate if between −0.5 and −1.0 dB/year; and fast if faster than −1.0 dB/year.¹¹

LMMs were also used to evaluate the effect of IOP as well the effect of age, race, gender, and baseline disease severity on rates of visual field loss. In addition, we also investigated the effect of type 2 diabetes and systemic arterial hypertension (SAH) on rates of change. These systemic age-related conditions are well known risk factors for microvascular disease and have been associated with other optic neuropathies.^{12, 13} Multivariable LMMs included baseline age, gender, race, follow-up time, IOP, and baseline disease severity (as defined by the MD value of the first valid SAP test for each eye). Separate multivariable models were built for mean IOP and peak IOP, as these variables were highly correlated.

All statistical analyses were completed in Stata (version 17, StataCorp LP, College Station, TX) within the Protected Analytics Computing Environment developed by Duke University for analysis of identifiable protected health information.

RESULTS

The study included 65 eyes of 43 patients with ODD, followed for a mean ± standard deviation (SD) of 7.6 ± 5.3 years (median 6.2, interquartile range [IQR] 3.4 to 11.0 years). Mean age at baseline was 55.8 ± 13.5 years; 29 (67.4%) subjects were female and 7 (16.3%) were self-identified as Black or African American. The mean MD at baseline was −4.42 ± 5.22 (median −2.69, interquartile range [IQR] −6.21 to −1.06 dB), ranging from −23.44 dB to 2.43 dB. Mean IOP during follow-up was 14.7 ± 2.7 mmHg and peak IOP was 17.0 ± 3.6 mmHg. Table 1 summarizes demographic and clinical characteristics of the eyes included in the study.

Table 1.

Demographic and clinical characteristics of eyes with optic disc drusen (ODD). Values represent mean ± standard deviation unless otherwise indicated.

Characteristic
Subject-specific	43 patients
Age, years	55.8 ± 13.5
Female, (%)	29 (67.4%)
Self-identified Race, (%)
Non-Black or African American	36 (83.7%)
Black or African American	7 (16.3%)
Eye-specific	65 eyes
Visual field Tests per eye, n*	4 (3, 5)
Follow-up, years	7.6 ± 5.3
Baseline MD, dB*	−2.69 (−6.21, −1.06)
Mean IOP, mmHg	14.7 ± 2.7
Peak IOP, mmHg	17.0 ± 3.6

IOP= intraocular pressure; IQR= interquartile range; SD= standard deviation.

Values are shown as mean±standard deviation unless otherwise noted.

^*Median (IQR)

Eyes with ODD had mean rate of SAP MD change of −0.23 ± 0.26 dB/year (median −0.16, [IQR] −0.25 to −0.08 dB/year). (Figure 1) Fifty-seven eyes (87.7%) had slow progression (slower than −0.5 dB/year), while 6 eyes (9.2%) had moderate progression (between −0.5 dB/year and −1dB/year), and only 2 eyes (3.1%) had fast progression (faster than −1 dB/year). Figure 2 shows representative examples of eyes with ODD, in which (A) represents one eye with slow rate of change, and (B) the eye with fast rate of change in SAP MD.

Figure 1.

(A) Distribution of rates of change in standard automated perimetry (SAP) mean deviation (MD) of eyes with optic disc drusen. (B) Percentage of eyes classified as slow, moderate, and fast progressors according to rates of change of SAP MD.

Figure 2.

Representative examples of eyes with optic disc drusen that showed (A) slow and (B) fast rate of standard automated perimetry (SAP) mean deviation (MD) over time. A1 and B1: Color fundus photos of the optic nerves; A2 and B2: Spectral Domain-Optical coherence tomography (SD-OCT) blue-light autofluorescence images of the optic nerves; A3 and B3: greyscale plots of baseline and follow-up visual field tests; A4 and B4: total deviation plots of baseline and follow-up visual field tests.

Univariable and multivariable regression models were used to investigate the association of clinical variables and IOP during follow-up with rates of SAP MD change over time (Table 2). In the multivariable models, older age and lower baseline MD were statistically significantly associated with faster rates of change, with 0.06 dB/year faster loss for each decade older (P= 0.044) and 0.03 dB/year faster loss for each 1 dB lower MD at baseline (P< 0.001). None of the IOP parameters were statistically significantly associated with rates of change in multivariable models. Figure 3 shows the relationship of IOP parameters with rates of SAP MD change. Type 2 diabetes and SAH were also not statistically significant in multivariable models.

Table 2.

Univariable and multivariable models investigating the effect of each clinical characteristic on the rate of change of standard automated perimetry (SAP) mean deviation (MD) over time in eyes with optic disc drusen (ODD).

	Univariable Models		Multivariable Model 1		Multivariable Model 2
Characteristic	Coefficient (95% CI)	P-Value	Coefficient (95% CI)	P-Value	Coefficient (95% CI)	P-Value
Age,
Per 10-year older	−0.108 (−0.164, −0.053)	<0.001	−0.060 (−0.115, −0.006)	0.030	−0.057 (−0.113, −0.002)	0.044
Gender,
Male	−0.352 (−0.193, −0.511)	<0.001	−0.105 (−0.056, 0.266)	0.203	−0.095 (−0.067, 0.256)	0.251
Race,
African American	−0.108 (−0.129, 0.345)	0.372	−.0.082 (−0.107, 0.271)	0.396	−0.080 (−0.110, 0.270)	0.409
Baseline MD,
Per 1dB lower	−0.034 (−0.019, −0.049)	<0.001	−0.028 (−0.013, −0.042)	<0.001	−0.026 (−0.012, - 0.041)	<0.001
Peak IOP,
Per 1 mmHg higher	0.004 (−0.018, 0.051)	0.704	0.028 (−0.025, 0.013)	0.558	---	---
Mean IOP,
Per 1 mmHg higher	0.024 (−0.004, 0.051)	0.096	---	---	0.002 (−0.022, 0.026)	0.884
Systemic Arterial Hypertension
Yes	−0.053 (−0.230, 0.123)	0.553	−0.029 (−0.176, 0.117)	0.695	−0.036 (−0.184, 0.112)	0.631
Diabetes Mellitus
Yes	0.012 (−0.229, 0.253)	0.920	−0.108 (−0.292, 0.077)	0.253	−0.118 (−0.304, 0.069)	0.217

CI = confidence interval; IOP= intraocular pressure; MD= mean deviation.

Figure 3.

Scatterplot of the relationship between (a) mean intraocular pressure (IOP) and (b) peak IOP, and rates of change in standard automated perimetry (SAP) mean deviation (MD) in eyes with optic disc drusen (ODD).

In our sample, 17 eyes (26.2%) were treated with IOP lowering drugs. These eyes had a mean MD at baseline of −7.27 ± 5.83 dB, while untreated eyes had a mean baseline MD of −3.41± 4.66 dB (median −6.21 dB [IQR: −9.00 to −2.53 dB], vs-1.70 dB [IQR: −4.71 to −0.69 dB], respectively; P<0.001). There was no significant difference in the mean rates of change between eyes receiving IOP lowering drugs and eyes that did not receive IOP lowering medication (−0.19 ± 0.24 dB/year vs. −0.23 ± 0.24 dB/year, respectively; P = 0.661). In multivariable models investigating the effect of mean and peak IOP on rates of SAP change in untreated eyes, none of these parameters had a significant association. (Supplemental Table 1)

According to clinicians’ notes, 15 eyes (23.1%) had a concurrent POAG diagnosis. There was no statistically significant difference in rates of MD change in ODD eyes that had a concomitant glaucoma diagnosis and those that did not (−0.18 ± 0.24 dB/year vs. −0.23 ± 0.24 dB/year, respectively; P= 0.585).

DISCUSSION

In this study, we assessed rates of visual field loss in eyes with ODD and evaluated the effect of clinical characteristics and IOP over time on those rates. We found that most eyes with ODD had slow rates of change in SAP MD. Older age and severity of visual field loss at baseline were significantly associated with faster rates of change over the study period. These findings may have significant implications for counseling patients diagnosed with ODD in terms of visual prognosis.

ODD can cause visual field defects varying from localized nerve fiber bundle defects to generalized constriction.¹⁴ Such defects can progress over time and cause vision loss.¹⁵ Lee and collaborators,³ have reported an average rate of loss of 1.58 ± 0.28% per year in a sample of 60 eyes with ODD using manual Goldman perimetry over an average follow-up period of 3 years. In our study, we evaluated rates of change using computerized perimetry that overcomes many limitations of Goldman perimetry, such as examiner bias, and lack of reproducibility and standardization of the examination. We found that eyes with ODD had an average rate of change of −0.23 ± 0.26 dB/year. Interestingly, this average rate of change is similar to those reported for POAG eyes in large clinical populations.^{7, 11} However, 8 (12.3%) ODD eyes in our sample presented rates of SAP loss that would be considered moderate or fast. Those eyes progressed at an average rate of −0.83 ± 0.61 dB/year. Given an average age at baseline of 55 years, it is likely that such rates would lead to substantial loss of vision over time in these patients if persistent. Therefore, there seems to be a small proportion of patients with ODD that present fast visual field loss which may end up leading to significant visual impairment in their lifetimes.

In our study, older age was significantly associated with faster rate of loss of SAP MD. It has been previously reported that as age increases, ODD may become more visible, and visual field defects more prevalent.^{15, 16} In the study by Lee et al., older age was significantly associated with lower visual field scores on Goldman perimetry. In our study, although a significant effect of age was noted, it seems to be of relatively small magnitude, with each decade older associated with 0.06 dB/year faster rate of loss (P=0.044).

Another important clinical question is if IOP could be related to visual field loss in eyes with ODD. It has been presumed that ODD decreases blood flow through perioptic nerve arterioles,¹⁷ a mechanism that has also been considered to be involved in the pathogenesis of glaucomatous optic neuropathy.¹⁸ Given this similarity, it has been thought that lowering IOP could improve perfusion pressure in optic nerves suffering by mechanic compression from drusen. Oliveira-Ferreira and collaborators⁴ reported a significant association between IOP and perimetric MD (Spearman’s rho = −0.863, P< 0.01) in a study including 36 eyes. This study is clearly limited by a cross-sectional design. In contrast, we investigated the effect of clinical IOP control on functional change longitudinally. Neither mean IOP nor peak IOP were associated with rates of SAP MD loss in univariable and multivariable models.

Of note, 17 eyes with ODD received IOP-lowering medication during follow-up. A previous study from Pojda-Wilczek⁶ evaluated 34 patients with bilateral ODD prospectively in which the eye with the more advanced disease was treated with brinzolamide. At the end of the 1-year follow-up they found a significant increase in MD in eyes that were treated with brinzolamide. However, it is possible that learning effects on perimetry may have confounded their results. In our study, there was no significant difference in rates of change between treated and untreated eyes (P=0.661). In addition, we investigated eyes that were left untreated during follow-up. For theses eyes, IOP parameters were also not significantly associated with rates of SAP MD change over time.

Our study had limitations. This was a retrospective study and patients were managed at the discretion of their attending ophthalmologists during follow-up. It is possible that characteristics of ODD eyes at presentation may have influenced clinicians’ decisions regarding when to perform visual fields and whether to prescribe IOP-lowering treatment, and this may have influenced the associations found in our study. However, the analyses of rates of change as performed in our study provides prognostic information on the clinical course of eyes with ODD that are managed according to current clinical practice. In addition, there was not a significant difference in rates of change between treated and untreated eyes. As another limitation, some eyes had few SAP tests acquired over time. It could be argued that the number may have been insufficient for a precise estimation of rates of change in some eyes. To assess that, we also conducted a subanalysis using only the eyes that had at least 5 visual fields over time and found an average rate of change of −0.23 ± 0.22 dB/year. Of note, we were not able to quantify the impact of the number and location of drusen in the optic nerve. It has been described that superficial drusen may cause a more prominent visual field defect compared with buried drusen.^{14, 19} However, we included eyes that had a broad range of SAP MD values at baseline which possibly accounted for different numbers and drusen locations.

In conclusion, we found that although most ODD eyes seem to progress slowly, around 12% of the eyes presented with moderate or fast rates of change that may incur significant visual field loss over time. Older age and severity of visual field loss at baseline were significantly associated with faster rates of change and should be considered for patient counseling regarding prognosis. Future studies should evaluate potential vascular and compressive mechanisms contributing to visual field loss in eyes with ODD.

Abbreviations:

CPT

current procedural terminology

DOR

Duke Ophthalmic Registry

EHR

electronic health records

GAT

Goldman applanation tonometry

HFA

Humphrey Field Analyzer

ICD

international classification of diseases

IOP

intraocular pressure

IQR

interquartile range

LMM

linear mixed model

MD

mean deviation

OCT

optical coherence tomography

ODD

optic disc drusen

PACE

protected analytics computing environment

POAG

primary open-angle glaucoma

SAH

systemic arterial hypertension

SAP

standard automated perimetry

SD

standard deviation

SITA

Swedish Interactive Threshold Algorithm

VIP

Vision, Imaging and Performance