Cupped disc with normal intraocular pressure: the long road to avoid misdiagnosis.

We present a series of six patients who had been receiving treatment for normal tension glaucoma (NTG; five patients) or primary open angle glaucoma (one patient). All of them were found to have optic neuropathy secondary to compression of the anterior visual pathway. Even though uncommon, compression of the anterior visual pathway is an important differential diagnosis of NTG. Diagnosis of NTG should be by exclusion. Here the possible causes of misdiagnosis are discussed. We present an approach to distinguish glaucomatous from nonglaucomatous optic neuropathy. The article also emphasizes how important it is for the clinicians to consider the total clinical picture, and not merely the optic disc morphology, to avoid the mismanagement of glaucoma, especially the NTG.

Glaucomatous optic neuropathy is characterized by progressive loss of the nerve fiber layer resulting in diffuse loss or notching of the neuroretinal rim especially to the optic disc margin.[1] Other characteristic features of glaucomatous optic neuropathy include optic disc hemorrhage crossing the neuroretinal rim, intereye asymmetry of cupping in the absence of asymmetry of disc size, and parapapillary atrophy.[1]

The current definition of glaucoma precludes intraocular pressure (IOP) as a defining feature.[2] One can diagnose glaucoma even when the IOP is “normal.” Unfortunately, in day-to-day practice, excessive cupping of the optic disc is considered to be pathognomonic of chronic glaucoma. One tends to diagnose “normal tension glaucoma” (NTG) if the IOP falls within the acceptable range often without investigating nonglaucomatous causes. In fact, in many cases, long-standing optic neuropathy can result in optic disc cupping. Causes of nonglaucomatous optic disc cupping include methanol poisoning,[3] arteritic anterior ischemic optic neuropathy,[4] and rarely chronic compressive lesions of the optic nerve.[5-8]

In the present series, we report six patients, who were misdiagnosed and treated as having glaucoma, in whom bilateral optic disc cupping simulating glaucomatous optic neuropathy was a feature of compression of the anterior visual pathway. We discuss an approach to the not so rare clinical presentation of cupping of the optic nerve head associated with normal IOP to prevent similar occurrences in future.

Case Report

The cases are summarized in Table 1.

Table 1

Details of cases

Discussion

The patients in this series ranged in age from 40 to 69 years. All of them had presented to their treating ophthalmologist with painless, gradually progressive vision loss. Except for a diffuse, dull headache at presentation to us in only one patient, none of them had any neurological symptoms. All were diagnosed to have glaucoma and were treated with IOP lowering measures. Three of them had undergone trabeculectomy. All, except the last, presented to us with progressive vision loss despite adequate treatment for glaucoma.

Pruett et al.[8] cited “a low index of suspicion” as a main culprit in the delayed diagnosis of chiasmal compression in 1973. This series illustrates why one should be extra-vigilant when diagnosing NTG even today.

All these patients had signs that should have raised suspicion about the presumptive diagnosis of NTG [Table 2]. Notably, the morphology of the pupils and that of the optic nerve head excepting the vertical cup-to-disc ratio were not recorded in any of these patients. Prominent pallor that was more than loss of the neuroretinal rim, at least when we examined, was seen in all, except the fifth patient. Misinterpretation of the visual fields is another significant factor for the misdiagnosis in these cases. For example, patchy visual field loss in the right eye of the second patient's earliest visual field [Fig. 6B] could be mistaken for glaucomatous biarcuate visual field loss. On careful evaluation, the visual field defects in the earliest visual field respected the vertical meridian. Thinning of retinal nerve fiber layer (RNFL) on optical coherence tomogram might have been considered to represent glaucomatous optic neuropathy in the fourth patient [Figure 10A]. Thinning of RNFL is the end result of any pathological process causing optic neuropathy and does not necessarily mean glaucoma. A comprehensive ophthalmic evaluation and appropriate interpretation of the clinical and/or investigational findings could have saved the second, third, and fifth patients from an erroneous diagnosis of glaucoma and glaucoma filtering surgeries. Moreover, neuroimaging at least when the visual loss was progressive despite achieving low IOP could have prevented further visual loss in them as well as in the first patient.

Table 2

Probable factors that lead to misdiagnosis in the presented cases

Figure 6

Case 2 – (A) The visual fi eld of the right eye in 2007. (B) The visual field of the right eye in 2002. Note the defects respect the vertical meridian

Figure 10

Case 4 – Optical coherence tomogram (A) and automated perimetry (B) in the left eye. Note the gray scale; the visual field defect respects the vertical meridian. The pattern deviation is misleading in advanced visual field loss

Case 1 – The earliest available visual fi eld in 2003; left eye (A) and right eye (B). Advanced fi eld loss in the left eye and superotemporal quadrantanopia in the right eye

Case 1 – Minimal pallor of the temporal rim in the right eye (A) and gross pallor of the rim in the left eye (B)

Case 1 – The visual field of the right eye at presentation to us in February 2008. Note the progression in the fi eld defect

Case 1 – Magnetic resonance imaging of brain (sagittal section) shows a well-circumscribed, lobulated sellar mass lesion suggestive of pituitary macroadenoma (outlined by arrows). Superiorly, it is displacing and compressing the optic chiasm

Case 2 – Thin and pale neuroretinal rim of both optic discs

Case 2 – Axial computed tomography scan shows a suprasellar mass (outlined by arrows) suggestive of pituitary adenoma

Case 3 – Area of pallor more than that of cupping in both optic discs

Case 3 – Coronal section on magnetic resonance imaging of brain shows a well-circumscribed, lobulated, homogeneous, solid suprasellar mass lesion consistent with meningioma (outlined by arrows). Intracranial optic nerves and optic chiasm could not be identified separately from the lesion

Case 4 – Magnetic resonance imaging of brain (sagittal section) shows a cystic lesion in the suprasellar cistern (outlined by arrows) suggestive of craniopharyngioma

Case 5 – Visual field in the left (A) and right (B) eye. Note gross intereye asymmetry in the visual field despite symmetrical optic disc cupping [Fig. 12]. The field defect in the left eye respects the vertical meridian

Figure 12

Case 5 – Optic disc photographs. Note the absence of pallor of the neuroretinal rim

Case 5 – Computed tomography scan of brain (axial section) shows a well-defined, cystic suprasellar lesion with calcifi cation (outlined by arrows) suggestive of craniopharyngioma

Case 6 – Right (A) and left (B) optic disc photographs. Note appreciable pallor only of the temporal rim in the left eye (arrow)

Case 6 – Asymmetric bi-superotemporal quadrantanopia. Note that the visual field respects the vertical meridian in both eyes

Case 6 – Magnetic resonance imaging of brain (sagittal section) revealing a pituitary adenoma

Distinguishing glaucomatous from nonglaucomatous optic disc cupping on clinical examination is difficult. Generally, the optic disc in patients with intracranial compressive lesions is pale and lacks cupping as seen in glaucoma.[10] However, this is not the rule. Kupersmith[7] demonstrated cupping of the optic nerve resembling glaucoma in 16 patients with lesions compressing the anterior visual pathway. Bianchi-Marzoli[11] demonstrated an increased median cup-to-disc area in patients with compressive lesions compared to age-matched controls. Similarly, pallor of the optic disc might also be lacking in intracranial compressive lesions (e.g., in the fifth case), especially in early compression. At presentation, unequivocal pallor of one or both optic discs was noted in only 28 of 50 (56%) patients with chromophobe pituitary adenoma causing visual field defects.[12] Pallor of the optic disc can also be masked by lenticular changes. Trobe[13] tested whether glaucomatous and nonglaucomatous optic disc cupping can be distinguished ophthalmoscopically. Two glaucoma specialists and one neuro-ophthalmologist independently analyzed optic disc stereo photographs. A total of 13 out of 29 (44%) eyes with nonglaucomatous optic neuropathy were classified as glaucomatous by at least one observer, demonstrating that even experts can misdiagnose glaucomatous cupping on isolated stereo photographs.

Visual field examination is another important factor to distinguish glaucomatous from nonglaucomatous optic neuropathy. However, since the arcuate distribution of the optic nerve fiber bundles remains intact as distal as the anterior chiasm, compression of the intracranial optic nerve or the anterior chiasm can produce arcuate visual field defects resembling glaucoma with sparing of central vision. Ahmed[14] prospectively performed neuroimaging in NTG patients who presented with classic glaucoma features without overt neurologic signs. Compression of the anterior visual pathway was found in 4 of 62 (6.5%) patients, all of whom had visual field defects typical of glaucoma.[14]

Routine neuroimaging for patients with presumed NTG is considered unnecessary due to a low yield for detecting intracranial pathology.[10] The South East Asia Glaucoma Interest Group (SEAGIG) recommends appropriate neuroradiological investigation only in a small proportion of patients with NTG, especially those who are younger and/or have unilateral disease, optic disc pallor out of proportion with cupping, atypical visual field defects, and color deficiency.[15] These factors, even though highly specific, have low sensitivity.[10] Greenfield compared eyes of 23 NTG patients with eyes with optic disc cupping associated with intracranial masses.[10] Age younger than 50 years, optic nerve pallor in excess of cupping, and vertically aligned visual field defects individually were only 46.4%, 45.5%, and 47.7%, respectively, sensitive for nonglaucomatous cupping.[10] One needs a highly sensitive examination or test to rule out nonglaucomatous cupping in patients with presumed NTG if a routine neuroradiological investigation is to be avoided.[16] In day-to-day clinical practice, we suggest combining the results of one or more factors listed in Table 3 to distinguish glaucomatous from nonglaucomatous cupping. If a patient lacks one characteristic of nonglaucomatous visual loss, there is a significant likelihood that another characteristic is present. For example, even though there was no neuroretinal rim pallor in either optic disc in the fifth patient, the nature of the visual field defect was a tipoff for the nonglaucomatous etiology. Performing a careful evaluation of the total clinical picture initially and maintaining a high index of suspicion during follow-up, while investigating appropriately if visual acuity, optic disc, or visual fields do not follow the expected pattern, we can minimize the chance of a potential diagnostic error.

Table 3

Distinguishing glaucomatous from nonglaucomatous optic neuropathy

In conclusion, it is important to remember that space occupying lesion in the parasellar and suprasellar region may be associated with contour abnormalities of the optic disc ophthalmoscopically indistinguishable from glaucomatous optic neuropathy. An accurate and early diagnosis can make a real difference to the lives of such patients by preventing visual impairment and significant morbidity. Similar to the article by Thomas,[17] this report points toward nonadherence to the routine comprehensive eye examination. This practice should change.