Paradoxical Changes of Retinal Nerve Fiber Layer Thickness in Uveitic Glaucoma

Importance  Measurement of retinal nerve fiber layer (RNFL) thickness using optical coherence tomography can aid in the diagnosis and management of glaucoma. We observed a previously unreported phenomenon in eyes with uveitis-associated glaucoma in which paradoxical changes in RNFL thickness were noted.

Observations  Four eyes of 3 patients with uveitis-associated glaucoma had a relatively normal RNFL measurement on presentation during periods of active uveitis and raised intraocular pressure. Subsequent control of uveitis and intraocular pressure was associated with a paradoxical thinning of the RNFL and increased cupping.

Conclusions and Relevance  Normal-appearing measurements of RNFL thickness in patients with uveitis should be interpreted cautiously in those with a raised intraocular pressure. Physicians should recognize that continued thinning of the RNFL and increased cupping, despite good intraocular pressure control in such eyes, may be due to resolution of edema of the RNFL.

Optical coherence tomography (OCT) is commonly used in glaucoma diagnosis and management. Thinning of the retinal nerve fiber layer (RNFL), measured using OCT, is indicative of glaucoma progression. Progressive glaucomatous change is also seen as increased cupping of the optic nerve. Three patients with uveitic glaucoma made us question this paradigm, and we urge physicians to use caution when relying on these modalities to evaluate such patients. This study was approved by the Duke University Institutional Review Board. A waiver of consent was obtained from all patients.

A 27-year-old man with bilateral uveitis secondary to Behcet disease was first evaluated in the glaucoma clinic 6 months after a fluocinolone acetonide–sustained drug delivery implant had been placed in both eyes. Visual acuity was 20/20 OU, intraocular pressure (IOP) was 22 mm Hg OU with 2 glaucoma eyedrops, and the slitlamp examination revealed quiet anterior chambers. He was bilaterally pseudophakic with symmetric cup-disc ratios (CDRs) of 0.5. Initial RNFL OCT was interpreted as normal in both eyes (Figure 1A). Six months later, IOP was 32 mm Hg OU, and the ophthalmic examination was otherwise unchanged. In both eyes, follow-up RNFL OCT demonstrated early inferotemporal thinning with mild superotemporal thickening (Figure 1B), and automated perimetry revealed new superior arcuate changes. An Ahmed glaucoma tube was placed sequentially in both eyes. Two months postoperatively, he had an IOP of 12 mm Hg OU with 2 glaucoma eyedrops, no uveitis activity, but an increased CDR of 0.8 in both eyes. Optical coherence tomography documented further RNFL thinning both superotemporally and inferotemporally (Figure 1C). Increased frequency of IOP checks revealed a stable IOP in the low teens at all visits. Three months later, IOP and clinical examination were stable, but overall severe progressive RNFL OCT thinning was noted (Figure 1D). Subsequent visits for the next 2 years, however, were unchanged, with a stable visual field, IOP, and RNFL.

A 73-year-old man with a visual acuity of 20/20 OS and an IOP of 24 mm Hg with no eyedrops in the left eye had been diagnosed with idiopathic unilateral uveitis. Slitlamp examination revealed low-grade anterior uveitis in the left eye with a CDR of 0.7 and an inferior notch. Baseline RNFL OCT demonstrated corresponding inferotemporal thinning (Figure 2A). The inflammation resolved with a several-month slow taper of topical corticosteroids while the IOP was treated with topical medications. One year later, he had an IOP of 35 mm Hg OS with 3 glaucoma eyedrops, an active low-grade anterior uveitis, and early posterior synechiae formation. The CDR was unchanged, and RNFL OCT demonstrated slight thickening compared with the previous year (Figure 2B). The patient’s inflammation and IOP responded to topical corticosteroids, and he achieved stability on a once-every-other-day corticosteroid regimen. At last follow-up 6 months later, he had an IOP of 11 mm Hg OS with 3 glaucoma eyedrops, with no uveitic activity detected. However, RNFL OCT demonstrated significant global thinning, despite a stable CDR, compared with the most recent scan as well as his baseline measurement (Figure 2C). Optical coherence tomography in the unaffected eye remained stable.

A 61-year-old woman with early open-angle glaucoma had a visual acuity of 20/20 OD and an IOP of 19 mm Hg OD with 1 glaucoma eyedrop. The CDR was 0.5, visual fields were full, and a baseline RNFL OCT was interpreted as normal (Figure 3A). A year later, the patient sought treatment for headaches. She had an IOP of 60 mm Hg OD. Slitlamp examination revealed low-grade anterior uveitis (diagnosed as idiopathic on investigation) and a CDR of 0.7 with inferior thinning. She started to receive topical corticosteroids, and the following day, her IOP was 19 mm Hg OD. On OCT, the RNFL was thickened (Figure 3B). A month later, she had decreased inflammation with topical corticosteroids but an IOP of 35 mm Hg OD with 3 glaucoma eyedrops. On OCT, the RNFL had mild inferior thinning (Figure 3C). A trabeculectomy with mitomycin C was performed. Her postoperative course was uncomplicated, and at a follow-up 2 months later, she had an IOP of 18 mm Hg OD with a once-daily topical corticosteroid, no uveitic activity, and a CDR of 0.8 with an inferior notch. On OCT, there was progressive superotemporal and inferotemporal RNFL thinning (Figure 3D) with further severe global RNFL thinning 5 months later (Figure 3E). Optical coherence tomography in the unaffected eye remained stable.

These cases highlight a previously unreported phenomenon in patients with uveitic glaucoma: “normal” RNFL measurements obtained during uncontrolled periods of uveitic activity with subsequent thinning following control of uveitis. These cases also demonstrate that despite apparent control of IOP and continued uveitis quiescence, the RNFL continued to thin and the optic nerve had increased cupping.

When glaucoma is superimposed in an eye with uveitis, a normal-appearing RNFL thickness^1-4 could lead a physician to be less aggressive in the management of IOP. Conversely, if a uveitis-associated normal-appearing RNFL becomes thinner due to uveitis control, it may appear as glaucoma progression. This finding may lead the physician to inappropriately conclude there has been an abrupt worsening of the glaucoma. When interpreting these measurements, physicians should recognize that normal values are for eyes without disease such as uveitis, thickness may vary with uveitic activity, and clinical judgment should be used to determine appropriate treatment.

Furthermore, on ophthalmoscopic evaluation, the optic nerve head may erroneously appear normal. We hypothesize that this normal-appearing optic nerve is likely due to subclinical RNFL edema,⁵ and after appropriate uveitis control, the edema resolves, the cupping increases and the RNFL thickness decreases, occasionally to the dramatic degree demonstrated earlier. Two patients showed an increase in RNFL thickness with increased inflammation, indicating that some of the thickness was secondary to edema. Furthermore, increased thickness of the RNFL noted after cataract or glaucoma surgery^6-8 could be due to mild uveitis that accompanies the postoperative course. Since none of the eyes demonstrated clinical edema, fluorescein angiograms were not obtained. It is possible that such testing may help determine if the normal-appearing nerve actually has some edema by revealing leakage. Although such a phenomenon could be due to resolution of RNFL edema, confirmation that a normal and stable IOP has been achieved is necessary in patients with uveitic glaucoma due to the labile nature of IOP control in such eyes.

Submitted for Publication: August 24, 2013; final revision received December 24, 2013; accepted December 29, 2013.

Corresponding Author: Sanjay Asrani, MD, Duke University Eye Center, PO Box 3802, Durham, NC 27710 (sanjay.asrani@duke.edu).

Published Online: May 22, 2014. doi:10.1001/jamaophthalmol.2014.954.

Author Contributions: Drs Asrani and Moore had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Asrani, Jaffe.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Asrani, Moore.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Moore.

Obtained funding: Jaffe.

Study supervision: Asrani, Jaffe.

Conflict of Interest Disclosures: None reported.