Transient Visual Loss in Young Females with Crowded Optic Discs: A Proposed Aetiology

ABSTRACT

I present four cases of transient visual loss (TVL) in young females with crowded optic discs. One patient had asymmetrical cup-to-disc ratios and only experienced TVL in the eye with the more crowded disc. I review the evidence for blood flow autoregulatory dysfunction within crowded optic discs in combination with reduced ocular perfusion pressure to propose a possible aetiology for both unilateral and bilateral TVL in young females with crowded optic discs.

Introduction

Transient visual loss (TVL) can be unilateral,¹ alternating unilateral² or bilateral and simultaneous.¹ The episodes of visual loss typically last from seconds to minutes but can persist for more than an hour before complete resolution.^3–5 The aetiologies of TVL have been previously divided into five groups: embolic; haemodynamic; vascular; optic disc/optic nerve/neurological; and miscellaneous.⁶ Emboli can originate from atherosclerotic disease, carotid dissection, or cardiac thromboembolism.⁶ Haemodynamic TVL is caused by non-embolic ocular hypoperfusion and can therefore be secondary to processes affecting systemic blood pressure, such as acute hypovolaemia.¹ Haemodynamic TVL is usually bilateral and simultaneous.¹ Vascular aetiologies include vasculitis and vasospasm, although retinal vasospasm can be a diagnosis of exclusion made without a reversible narrowing of the retinal vessels being visualised.^6,7 The subgroup of ‘optic disc/optic nerve/neurological’ includes papilloedema, optic disc drusen, migraine, and epilepsy.⁶ The final category of “miscellaneous” includes glaucoma and idiopathic cases.⁶

Reports of TVL in patients under the age of 40 suggest that identifiable embolic aetiologies are rare in this group with the majority of patients having a good prognosis following a TVL episode.^3–5 A significant proportion of TVL cases in younger patients can therefore be classified as idiopathic.⁵

I present four cases of TVL in young females with crowded optic discs. In order to explain the recurrent TVL episodes in this cohort, I review the evidence for blood flow autoregulatory dysfunction within crowded optic discs in combination with reduced ocular perfusion pressure (OPP) and propose a possible aetiology for both unilateral and bilateral TVL in young females with crowded optic discs.

This case series follows the tenets of the Declaration of Helsinki and has fulfiled the criteria required by the regional Quality Improvement Team, Research and Development Department, and Research Ethics Committee.

Report of cases

My cohort consists of four females (age range at presentation 14–22 years: see Table 1 for demographics) with crowded optic discs, defined as a cup-to-disc ratio of 0.2 or less.² Patients 1 to 3 experienced bilateral and simultaneous TVL. Patient 4, with asymmetrical cup-to-disc ratios and a more crowded optic disc in the left eye, experienced monocular TVL in the left eye only. The patients presented between 2008 and 2017. The specific phenomenologies of the TVL episodes for each patient are summarised in Table 2.

Table 1.

Demographics, investigations, medical and social histories for all patients

Patient number	Age at presentation	Visual acuities	Investigations to exclude optic disc drusen	Additional investigations	Associated medical and social histories
1	22	Right: 6/5 Left: 6/5	Fundal examination followed by repeat fundal examination 11 years after initial presentation	Normal MRI head/ECG/tilt table test	Smoker
2	20	Right: 6/6 Left: 6/6	Optic disc B-scan ultrasonography Optic disc OCT	Normal CT head/MRI head/MRV 1st LPOP 26 cmCSF, 2nd LPOP 20 cmCSF; both with normal constituents	Nil
3	14	Right: 6/6 Left: 6/6	Optic disc B-scan ultrasonography Optic disc OCT Optic disc fluorescein angiography Optic disc autofluorescence imaging	Craniopharyngioma diagnosed on MRI head having presented with a 3 month history of TVL, one syncopal episode whilst standing and amenorrhoea for 6 months Elevated optic discs noted at initial presentation – 1st LPOP “normal,” 2nd LPOP 10.5 cmCSF; both with normal constituents 24 hour ambulatory BP monitoring: see text	Craniopharyngioma contacting but not compressing the optic chiasm – excised and prescribed hydrocortisone and levothyroxine for hypopituitarism TVL continued after surgery (normal cortisol levels on three separate occasions)
4	18	Right: 6/6 Left: 6/36*	Optic disc B-scan ultrasonography Optic disc OCT Optic disc fluorescein angiography Optic disc autofluorescence imaging	ECG, EEG, baseline blood tests normal (investigations for previous syncopal attacks) 24 hour ambulatory BP monitoring – no TVL episodes during monitoring period	History of five syncopal attacks whilst standing Supine BP 133/72 mmHg, standing BP 114/63 mmHg – insufficient for a diagnosis of orthostatic hypotension**

BP = blood pressure; CSF = cerebrospinal fluid; CT = computerised tomography; ECG = electrocardiogram; EEG = electroencephalogram; LPOP = lumbar puncture opening pressure; MRI = magnetic resonance imaging; MRV = magnetic resonance venography; OCT = optical coherence tomography; TVL = transient visual loss

* Amblyopic left eye

** Orthostatic hypotension is defined as a systolic drop of at least 20 mmHg or a diastolic drop of at least 10 mmHg after 3 minutes of standing

Table 2.

Phenomenology of transient visual loss for each patient in the cohort

Patient number	Frequency	Duration of visual loss	Other features	Predisposing factors	Associated systemic symptoms
1	1 to 5 per day	20–30 seconds	Bilateral and simultaneous Visual fields constricting from the periphery towards fixation; vision was lost “to blackness” in both eyes	Only when upright, associated with a change in posture to standing	Frontal pressure-like headaches following the TVL episodes without hemicranial features, nausea or photophobia Intermittent episodes of disequilibrium, which were separate in time to the TVL episodes
2	10 per day	5 seconds	Bilateral and simultaneous	Only when upright	Frontal pressure-like headaches following the TVL episodes without hemicranial features, nausea or photophobia
3	3 to 10 per day	2 minutes	Bilateral and simultaneous	Only when upright, associated with a change in posture to standing	Nil reported
4	1 per day	Loss of vision to no perception of light over 5 seconds with recovery over 30 minutes	In left eye only (eye with the smaller cup-to-disc ratio)	No association between TVL and posture noted but history of syncopal episodes when upright	Generalised headaches following the TVL episodes without hemicranial features, nausea or photophobia. No response to sumatriptan

TVL = transient visual loss

Ocular examinations were normal other than for bilateral crowded optic discs (Figure 1). The left eye of Patient 4 with the more crowded optic disc (Figure 1d) was also amblyopic with a Snellen visual acuity of 6/36. Intraocular pressures (IOP) were 21 mmHg or less in all eyes from all of the patients. There was no evidence of optic disc drusen in any of the patients (see Table 1).

Figure 1.

Optic disc images for (a) Patient 1, (b) Patient 2, (c) Patient 3, and (d) Patient 4. (d) Optical coherence tomography sections through the optic discs of Patient 4 demonstrate the greater degree of crowding of the left optic disc compared with the right optic disc

Of note is Patient 3 whose TVL continued following surgical excision of her craniopharyngioma and despite normal cortisol levels on three separate occasions post-operatively. Therefore, 24-hour arterial BP monitoring with event recording was arranged. She experienced a single episode of bilateral simultaneous TVL whilst attached to the BP monitor, at 20:15. Although the monitor was not triggered at the time of the event, the monitor had recorded a BP before the TVL episode at 19:28, which was 106/74 mmHg, and after the TVL episode at 20:25 of 100/61 mmHg. The significance of these findings is discussed below.

Patients 1 and 3 were discussed with neurology regarding the possibility of the TVL being secondary to a hypoperfusive phenomenon (suggested by the bilaterality and temporal relationship to changing posture). For both patients, it was agreed to initiate a trial of fludrocortisone to assess if iatrogenically raising the BP would reduce the frequency of TVL episodes. Both Patients 1 and 3 were prescribed fludrocortisone 100 μg twice daily, which led to a reduction in the frequency of TVL, from between one and five episodes per day to one episode every 2 weeks for Patient 1 and from between 3 and 10 episodes per day to one episode per day for Patient 3. Fludrocortisone also reduced the frequency of the episodes of disequilibrium for Patient 1. BP on treatment for Patient 1 was 110/63 mmHg supine and 117/77 mmHg standing. The presenting sitting BP was 108/81 mmHg. The improvement in TVL frequency was sustained in Patient 1 but not in Patient 3 for whom the episodes of TVL returned to three episodes per day after 4 weeks of treatment. At the time of TVL recurrence, BP on treatment was 105/66 mmHg, and she elected to stop fludrocortisone.

Patients 1 and 2 defaulted from further follow up. The episodes of TVL experienced by Patient 3 spontaneously remitted at the age of 16 with a BP of 111/66 mmHg off fludrocortisone. The TVL episodes experienced by Patient 4 remitted at age 19 at which point she was discharged from the eye service. All patients were recontacted and surveyed in 2020 with the time of contact being from 3 to 12 years since the first presentation. None of the patients reported developing significant cardiovascular disease or any other new significant ophthalmic or medical diagnoses since discharge.

Discussion

As discussed, the five subclassifications of TVL are embolic, haemodynamic, vascular, optic disc/optic nerve/neurological, and miscellaneous (including idiopathic).⁶ None of our cohort had latterly developed significant cardiovascular disease, rendering embolic aetiologies less likely. In addition, none of the patients had been diagnosed with vasculitis, epilepsy, papilloedema, disc drusen, or glaucoma; diagnoses from the “vascular,” “optic disc/optic nerve/neurological,” and “miscellaneous” categories.⁶ Bilateral simultaneous TVL in three of the four patients would be inconsistent with retinal vasospasm, which is by definition monocular.⁷ The most likely aetiology, based on the patients’ symptoms, is therefore haemodynamic, which is usually bilateral and simultaneous.¹ In support of a haemodynamic aetiology, Ewing suggested that the evolution of field loss in haemodynamic TVL can differ from embolic TVL.⁸ Whereas embolic TVL is associated with an advancing “window blind” type of field loss, haemodynamic TVL can be associated with an “iris diaphragm” pattern of loss or progressive circumferential constriction of the peripheral field with the central vision being the last to be lost.⁸ The “iris-diaphragm” progression is similar to the symptoms described by Patient 1 (Table 2).

Haemodynamic aetiologies are defined by their association with ocular perfusion pressure or the difference between the arterial BP and the intraocular venous pressure.^9,10 Since the IOP is physiologically related and close in value to intraocular venous pressure,^11,12 IOP is used as a surrogate measure of intraocular venous pressure, which produces the formula below to calculate the Mean OPP, assuming an upright patient.⁹

Mean OPP = 2/3 MAP – IOP

Where MAP (mean arterial pressure) = diastolic BP + 1/3 (systolic BP – diastolic BP)

The 2/3 correction factor for MAP allows for the difference between brachial artery pressure and ophthalmic artery pressure in a seated patient.¹⁰

In addition to OPP tissue perfusion is also dependent on autoregulation, which maintains a steady blood flow within a defined band of perfusion pressures represented by the autoregulatory plateau. Autoregulation has been demonstrated to act within the optic nerve heads of humans.^13–18 Considering the lower end of the autoregulation plateau in humans, autoregulation becomes dysfunctional at an OPP of around 20 mmHg.^16,17,19 Using the OPP equation, the OPPs for Patient 3 around the time of her TVL episodes were at the lower end of normal for her age, sex, and height (BP before TVL: systolic 32^nd percentile, diastolic 74^th percentile; BP after TVL: systolic 15^th percentile, diastolic 33^rd percentile)²⁰ but are insufficient of themselves to reduce her ocular blood flow to beyond the lower break of the ocular autoregulatory plateau. This is important since ocular hypoperfusion in isolation is the conventional explanation for bilateral simultaneous hypoperfusive TVL episodes.¹ The TVL in Patient 3 can therefore not be explained by ocular hypoperfusion alone.

However, perfusion of the optic nerve head is also dependent on a normally operating autoregulatory response. Unstable optic nerve head autoregulation could combine with a low normal OPP, the result being optic nerve head hypoperfusion and TVL.²¹ The physiological evidence for the possibility of dysfunctional optic nerve head autoregulation contributing to the TVL in our cohort, in combination with low normal BP, is presented in Table 3. Dysfunctional optic nerve autoregulation has been demonstrated in healthy human subjects. Young females have been found to demonstrate a greater latency within the cerebral dynamic autoregulatory response compared with men (the dynamic phase of autoregulation being the initial autoregulatory response to a drop in OPP), and the dynamic phase of autoregulation is more closely related to neuronal activity than steady state static autoregulation (see Table 3).^{13,15,22–26} The dynamic autoregulatory response could therefore be particularly pertinent to the manifestation of TVL, especially TVL temporally associated with a change in posture.

Table 3.

Categorising the published evidence in support of autoregulatory impairment in association with crowded optic discs contributing to the aetiology of transient visual loss in our cohort of young females

	Principal findings
Dysfunctional optic nerve head autoregulation in healthy human subjects
	A significant variation in optic nerve head autoregulatory efficiency is demonstrated between normal subjects13
	A subgroup of young healthy individuals demonstrate abnormal optic nerve head autoregulation15
	Fluctuations in OPP, which should be within the normal autoregulatory plateau, can lead to significant dips in optic nerve head blood flow in healthy subjects with dysfunctional optic nerve head autoregulation22
Dysfunctional autoregulation in young healthy females
	Young females demonstrate a greater latency within the cerebral dynamic autoregulatory response compared with males23
Importance of dynamic autoregulatory dysfunction for the symptomatology of our cohort
	The dynamic phase of autoregulation is more closely associated with neuronal activity than static autoregulation and provides a more sensitive assessment of autoregulatory dysfunction within the optic nerve head24,25
	Pre-existing dynamic autoregulatory dysfunction could predispose patients to ischaemic events26
	Orthostatic hypotensive symptomatology reported to be more common in females although, in a separate study, episodes of orthostatic hypotension were equally common in male and female subjects. Possibility of dysfunctional dynamic autoregulation increasing the probability of manifesting symptoms of orthostatic hypotension, including TVL23,27
Role of autoregulatory dysfunction in the aetiology of NA-AION, a condition associated with crowded optic discs
	Article reviews histological studies suggesting no robust evidence to support either an embolic or thrombotic aetiology for NA-AION28
	32% of NA-AION cohort without hypertension, diabetes mellitus, cardiovascular disease, previous cerebrovascular accident, or migraine29
	In patients with NA-AION and associated medical conditions, one of the commonest conditions is diabetes mellitus, which is associated with vasculopathic risk but is also known to impair autoregulatory function25,30
	Hayreh at al. proposed an aetiology for NA-AION based on an upshift of the autoregulatory plateau in hypertensives31,32
Association between young females and NA-AION
	In subgroup analysis of NA-AION patients under the age of 40, females were affected more frequently than males30
Previously reported associations between anomalous optic discs and monocular TVL
	Retrospective case series of 4 patients with anomalous optic discs reporting monocular TVL precipitated by a change in posture (2 unilateral non-papilloedematous optic disc swelling, 1 optic disc drusen and a 42-year-old female with a unilateral anomalous optic disc). A combined mechanism was proposed with reduced local optic nerve head perfusion combined with systemic hypotensive episodes to explain the relationship between TVL and a change in posture21
	Retrospective case series of 29 patients (mean age 45.5 years, 17% hypertensive, 14% hyperlipidaemic, 10% diabetic) who presented with monocular TVL on waking. 90% were female and 48% had crowded optic discs. One patient had a unilateral crowded optic disc and experienced TVL only in the eye with the crowded optic disc. The authors hypothesised autoregulatory failure within the optic nerve head leading to a failure of function at low light levels2

NA-AION = non-arteritic anterior ischaemic optic neuropathy; OPP = ocular perfusion pressure; TVL = transient visual loss

Crowded optic discs are recognised as an important risk factor for non-arteritic anterior ischaemic optic neuropathy (NA-AION),²⁸ the commonest ischaemic optic neuropathy.³³ What is the evidence for NA-AION being secondary in part to autoregulatory dysfunction, to support an association between autoregulatory dysfunction and crowded optic nerves and so support the role for the crowded optic discs in the aetiology of our cohort’s TVL? Current theories for the aetiology of NA-AION are based on an impairment of axoplasmic flow within the crowded optic nerve head with secondary swelling.^28,34 The increased tissue pressure compresses the microvasculature causing a chronic hypoperfusion of the optic nerve head,³⁵ a type of compartmentalisation,²¹ which is not corrected by autoregulation.

Pertinent studies in support of the role of dysfunctional autoregulation in the aetiology of NA-AION are again summarised in Table 3. There is an absence of evidence for either an embolic or thrombotic aetiology for NA-AION and a significant association of NA-AION with diabetes mellitus, which is associated with vasculopathic risk but is also known to impair autoregulatory function.^25,28–30 In addition, the theory of nocturnal hypoperfusion of the optic nerve head in arterial hypertensives purported by Hayreh et al. as a trigger for NA-AION is dependent on dysfunction of optic nerve head autoregulation.^31,32

I have to explain the mechanism by which fludrocortisone reduced the frequency of the TVL episodes in Patients 1 and 3. If optic nerve head autoregulation is fully functional then both the pre- and post-treatment BPs of Patient 1 should maintain OPPs within the normal autoregulatory plateau. The effect of fludrocortisone can therefore not be explained by an iatrogenic increase in BP alone if associated with normal autoregulation. I note again studies demonstrating impaired autoregulation in a subgroup of healthy individuals,^13,22 specifically a linear relationship between increasing OPP and tissue blood flow.¹³ This effectively suggests zero functioning of optic nerve head autoregulation in some subjects. If Patient 1 had autoregulatory dysfunction to this degree then a small increase in OPP could manifest as an augmented improvement in optic nerve head perfusion, explaining the benefit of fludrocortisone. The improvement in symptomatology with fludrocortisone could therefore also be considered evidence for autoregulatory dysfunction.

In 2018, Bouffard et al. published a retrospective case series of 29 patients (mean age 45.5 years) who presented with monocular TVL on waking, of whom 90% were female and 48% had crowded optic discs.² Four patients had asymmetrical cup-to-disc ratios but only one patient had a crowded optic disc in the eye with the smaller cup-to-disc ratio and this patient experienced TVL only in the eye with the more crowded optic disc.² The authors hypothesised vascular autoregulatory failure within the optic nerve head, leading to a failure of function at low light levels. My case series has similarities to that of Bouffard et al. Both are retrospective, so investigations and interventions differ between patients. The majority of the patients of Bouffard et al. were female and my cohort is entirely female. Also, the fact that the only patient with asymmetrical cup-to-disc ratios experienced TVL in the eye with the smaller cup-to-disc ratio is a further similarity. My cohort, however, was younger and had no additional vascular risk factors (in the cohort of Bouffard et al. 17% were hypertensive, 14% had hyperlipidaemia and 10% were diabetic).²

A further difference between the series of Bouffard et al. and mine is that three of our four patients experienced bilateral simultaneous TVL whereas with the patients of Bouffard et al., the TVL episodes were exclusively monocular, although the side changed between attacks in 24% of their cohort.² Whereas the monocular TVL episodes in the series of Bouffard et al. were attributed to autoregulatory dysfunction alone leading to local optic nerve head hypoperfusion unilaterally,² my cases reporting binocular simultaneous TVL suggest an additional systemic hypoperfusive component. I would suggest therefore that our cohort of young females are demonstrating episodes of low normal BP insufficient in isolation to impair optic nerve head perfusion, but when combined with dysfunctional optic nerve autoregulation in the context of crowded optic discs, could have a synergistic effect on the impairment of optic nerve head perfusion.

I present a cohort of four young females with crowded optic discs, three with bilateral crowded optic discs reporting bilateral TVL and one patient with asymmetrical crowding reporting TVL only in the eye with the more crowded optic disc. I refer to studies designed to define the lower break of the ocular autoregulatory plateau in humans to confirm that there is insufficient evidence for systemic hypotension alone to explain the TVL in our cohort despite the TVL episodes being bilateral and simultaneous in three of the subjects.^16,17,19 I therefore propose a synergistic combination of low normal BP and dysfunctional autoregulation within crowded optic discs as a possible explanation for the TVL experienced by our patients. I cite studies suggesting dysfunctional optic nerve autoregulation in otherwise healthy individuals and a possible predisposition to dynamic autoregulatory dysfunction in females in support.^13,15,22,23 I cite evidence for dysfunctional autoregulation within crowded optic discs contributing to the association between optic disc crowding and NA-AION.^25,28,31,32 I note conditions such as open-angle glaucoma in which both autoregulatory dysfunction and OPPs that are low but still within the plateau for normal autoregulation are proposed as sufficient contributory factors for progressive optic nerve neural loss.^36–39 In addition, Patient 4 suffered recurrent syncopal episodes and associations have been reported between a syncopal predisposition and autoregulatory dysfunction.⁴⁰ I also note the older cohort of patients experiencing monocular TVL on wakening published by Bouffard et al.² of whom 90% were female, 48% had crowded discs and one patient who demonstrated asymmetrical cup-to-disc ratios, as in my cohort, only experienced TVL in the eye with a smaller cup-to-disc ratio.

I suggest therefore that crowded optic discs with dysfunctional autoregulation in combination with low normal BPs could be considered in younger female patients as a differential diagnosis for both binocular and monocular transient visual loss. My hypothesis remains dependent however on future technologies allowing improved quantitative assessment of optic nerve head blood flow since currently there is no quantitative method for assessing optic nerve head blood flow as an absolute value in the more posterior part of the optic nerve supplied by the short posterior ciliary arteries.¹⁰ If a physiological relationship is confirmed between dysfunctional autoregulation within crowded optic discs and TVL then this association could be a useful addition to the list of possible aetiologies for TVL in young females, especially in cases with asymmetric optic disc crowding and associated monocular TVL since other proposed aetiologies for monocular TVL in young patients in the absence of gross ocular or systemic pathology include conditions such as retinal vasospasm, which are often diagnoses of exclusion.⁷

Patient consent and ethics statement

Consent for publication has been granted by all subjects.

Declaration of interest statement

The author reports no conflict of interest.