Detection of Asymmetric Glaucomatous Damage Using Automated Pupillography, the Swinging Flashlight Method and the Magnified-Assisted Swinging Flashlight Method

Detection of Asymmetric Glaucomatous Damage Using Automated Pupillography, the Swinging Flashlight Method and the Magnified-Assisted Swinging Flashlight Method

Authors:
M. Waisbourd, B. Lee, M.H. Ali, L. Lu, P. Martinez, B. Faria, A. Williams, M.R. Moster, L.J. Katz, and G.L. Spaeth
Institution: Glaucoma Research Center, Wills Eye Hospital, Philadelphia, PA, USA
Correspondence: M. Waisbourd, Wills Eye Hospital, 840 Walnut Street, Philadelphia, PA 19107, USA
Email: mwaisbourd@willseye.org

Published in: Eye (2015) 29, 1321–1328
DOI: 10.1038/eye.2015.106
Published online: 26 June 2015

---

Abstract

Purpose:
To determine the sensitivity and specificity of various methods of detecting a relative afferent pupillary defect (RAPD) in patients with glaucoma-related diagnoses.

Patients and Methods:
Patients underwent RAPD evaluation using the swinging flashlight method (SFM), the magnifier-assisted SFM (MA-SFM), and automated pupillography using the Konan RAPDx. The main outcome measures were the sensitivity and specificity of these three methods in detecting visual field mean deviation (MD), cup-to-disc ratio (CDR), disc damage likelihood scale (DDLS), and retinal nerve fiber layer (RNFL) asymmetry.

Results:
Eighty-one patients were enrolled (60 with glaucoma, 21 with ocular hypertension or glaucoma suspect).

• 31% had MD asymmetry ≥5 dB

• 19.7% had CDR asymmetry ≥0.20

• 26.7% had DDLS asymmetry ≥2

• 38.2% had RNFL asymmetry ≥10 µm

Sensitivity of pupillography:

• MD asymmetry: 93.3% (95% CI 68.1–99.8)

• CDR asymmetry: 80.0% (95% CI 51.9–95.7)

• DDLS asymmetry: 100.0% (95% CI 73.5–100.0)

• RNFL asymmetry: 69.2% (95% CI 38.6–90.9)

Specificity ranged from 32.8% to 42.9% across variables.
Pupillography amplitude score correlated with MD asymmetry (r² = 0.41, p < 0.001); area under the ROC curve was 0.84.

Conclusion:
Automated pupillography showed higher sensitivity but lower specificity than manual methods in detecting asymmetric glaucomatous damage. Within this cohort, its case-finding ability was limited.

---

Introduction

Testing for a relative afferent pupillary defect (RAPD) is essential in evaluating glaucoma, as it reflects asymmetric optic nerve damage. RAPD correlates with visual field loss, optic disc damage, and thinning of the retinal nerve fiber layer (RNFL), and may even precede detectable field changes.

The traditional Swinging Flashlight Method (SFM), described by Levatin (1959) and refined by Thompson, remains a qualitative assessment that can be affected by factors like anisocoria or improper light positioning. The Magnifier-Assisted SFM (MA-SFM) uses a +20 D lens to improve visualization and sensitivity.

Automated pupillography, using devices like the Konan RAPDx, objectively measures amplitude, latency, and velocity of pupil constriction through infrared recording, offering quantifiable data for detecting asymmetric glaucomatous damage.

---

Materials and Methods

Study Design and Participants

A prospective cross-sectional study was conducted at Wills Eye Hospital (Philadelphia, USA).
Inclusion: Glaucoma, glaucoma suspect, or ocular hypertension (IOP >21 mmHg).
Exclusion: Conditions preventing binocular fixation or pupil visualization, recent ocular surgery (<4 weeks), or non-glaucomatous optic neuropathy.

Clinical Examination and Imaging

All participants underwent:

• Visual acuity and intraocular pressure measurement (Goldmann tonometry)

• Cup-to-disc ratio (CDR) and Disc Damage Likelihood Scale (DDLS) scoring

• Visual field testing (Octopus 900, Haag-Streit)

• Optical coherence tomography (Cirrus OCT, Zeiss Meditec)

Asymmetry thresholds:

• MD ≥5 dB

• CDR ≥0.20

• DDLS ≥2

• RNFL ≥10 µm

RAPD Testing Procedures

1. Swinging Flashlight Method (SFM):
Performed with a Finoff halogen transilluminator at full intensity, alternately illuminating each eye for 3 seconds. RAPD was identified by pupil dilation upon illumination.

2. Magnifier-Assisted SFM (MA-SFM):
Same procedure as above, but with a +20 D lens to magnify pupil observation.

3. Automated Pupillography (Konan RAPDx):
Used binocular infrared cameras to measure amplitude and latency of constriction to standardized light stimuli.
RAPD score = 10 × log10(OD/OS).
An RAPD was defined as amplitude ≥0.20 or ≤−0.20.

Statistical Analysis

Data were analyzed using SAS software. Sensitivity, specificity, correlations, and ROC curves were generated for each test method relative to asymmetry thresholds in MD, CDR, DDLS, and RNFL.

---

Results

Study Population

81 patients were included (56% female, mean age 66.2 ± 13.6 years).

• 74% had glaucoma

• 26% had ocular hypertension or were suspects

Asymmetry rates:

• MD ≥5 dB in 30.6%

• CDR ≥0.20 in 19.7%

• DDLS ≥2 in 26.7%

• RNFL ≥10 µm in 38.2%

Sensitivity and Specificity

Specificity ranged from 32–43% for pupillography, higher for manual tests.

Correlations

Pupillography amplitude score showed:

• r² = 0.41 (p<0.001) for MD

• r² = 0.27 (p<0.005) for RNFL

AUC values:

• MD asymmetry ≥5 dB → 0.84

• CDR ≥0.20 → 0.62

• DDLS ≥2 → 0.81

• RNFL ≥10 µm → 0.57

---

Discussion

This study confirms that automated pupillography detects asymmetric glaucomatous damage with higher sensitivity but lower specificity than manual methods.

MA-SFM improved upon the standard SFM by enhancing visualization, doubling RAPD detection rate. Pupillography correlated moderately with both functional (MD) and structural (RNFL) asymmetries.

These results align with prior findings showing that RAPD can precede field loss and that automated quantification provides valuable diagnostic information. However, the relatively low specificity suggests the current amplitude cut-off (0.2) may detect subtle or preclinical asymmetry not yet evident in structural metrics.

Limitations include small sample size and heterogeneous disease stages. Further research should refine pupillographic parameters and explore its role in glaucoma screening.

---

Summary

What Was Known Before:

• Automated pupillography can detect asymmetric glaucomatous damage.

• Few studies directly compared it with MA-SFM using both structural and functional parameters.

What This Study Adds:

• Direct comparison among SFM, MA-SFM, and pupillography.

• Pupillography had higher sensitivity and lower specificity than manual methods for detecting asymmetric glaucoma.

---

Acknowledgements

The authors thank Lalita Gupta, Camila Zangalli, Alice Zhang, Eileen Wang, and Michael Pro for their assistance.
Funding: Glaucoma Service Foundation to Prevent Blindness.

---

References

(Selected key citations retained as per original article)

1. Kaback MB et al. Am J Ophthalmol. 1976;81:462–468.

2. Johnson LN et al. Ophthalmology. 1988;95:1649–1655.

3. Lankaranian D et al. Trans Am Ophthalmol Soc. 2005;103:200–207.

4. Tatsumi Y et al. Br J Ophthalmol. 2007;91:633–637.

5. Tatham AJ et al. Ophthalmology. 2014;121:1185–1193.

6. Sarezky D et al. J Glaucoma. 2014;23:419–423.

7. Cook C et al. J Glaucoma. 2009;18:557–562.

---

Baixe aqui o artigo na integra