Accessory iris membranes: Removing the veil.

PURPOSE: To report a case of a young adult with dense bilateral accessory iris membranes (AIMs). OBSERVATIONS: AIMs can influence vision by multiple mechanisms. We discuss clinical examination and imaging considerations that can help parse optical and refractive complications to better guide intervention discussions. We also describe our surgical approach and perioperative management to help minimize trauma to the eye and maximize favorable surgical outcomes in these cases. CONCLUSIONS AND IMPORTANCE: This case highlights the excellent symptomatic, visual acuity and stereopsis gains that can be achieved following surgical intervention for this clinical entity, even in older patients.

Introduction

Up to 95% of newborns display incomplete involution of the anterior tunica vasculosa lentis.,2 These are usually diaphanous, do not affect acuity, and regress by 1-year of age., Rarely, bilateral dense membranes covering much of the pupil remain after birth and fail to involute. These may represent a clinically-distinct entity caused by hyperplasia of the superficial mesenchymal layer of the anterior iris stroma, termed accessory iris membranes (AIMs)., They are usually sporadic, but autosomal dominant inheritance can occur.,

AIMs may influence vision through multiple mechanisms that can often be parsed. Deprivation amblyopia is nonprogressive in adulthood and typically unaffected by dilation due to drops or changes in ambient lighting. Severity is often modest, as small openings transmit light to the developing retina. Small apertures may even improve acuity by ameliorating uncorrected refractive error, either equally in all meridians via a pinhole effect when they are round, or preferentially in all-but-one meridian via a stenopaic effect when they are ovoid; these effects too should not change over time, but may diminish with dilation. Excessively small apertures may impair acuity by diffraction, chromatic aberration and/or reduced retinal luminance,,; these too should not fluctuate over time, but in contrast may improve with dilation. Inadequate retinoscopic reflex and/or subjective refinement may manifest as reduced acuity that uniquely fluctuates over time. Lenticular-AIM attachments can induce cataractous changes that may be difficult to observe clinically with dense AIMs but may uniquely diminish acuity over time. Finally, closely-approximated AIM apertures can cause monocular diplopia that varies with working distance due to the Scheiner principle. Several factors can exist in a single patient, making careful examination and imaging essential to optimize visual acuity and guide intervention. Observation may be appropriate if AIMs are bilateral and symmetrical, apertures are adequate in size and location, and a fixation preference is absent in pre-verbal patients. Refractive correction and (precautionary) patching may need to be employed. Treatment options include pharmacological mydriasis, surgical membranectomy or laser membranotomy, each of which offer advantages and disadvantages.,,

Case report

A 20-year-old Asian male college student majoring in computer science, presented with a two-year progressive subjective decline in vision which was unimproved by new glasses, and visual confusion (seeing different images in each eye) at near but not distance. He also endorsed nyctalopia and difficulty adjusting to changes in ambient lighting, but no hemeralopia or photoaversion. His parents had also noted his use of an abnormal head posture during computer use which he felt helped alleviate the visual confusion. Medical and family histories were noncontributory. He was originally evaluated elsewhere at three- and six-months of age, but given good monocular tracking and no fixation preference, observation was advised. His only consequent intervention was full myopic refractive correction from 8-years of age. On examination, best-corrected visual acuity (BCVA) was 20/40−2 in the right eye and 20/50+2 in the left eye under ambient lighting conditions, with no measurable objective change under scotopic or photopic conditions, or following pharmacological dilation. Alignment was orthophoric, with full eye movements, fusion at both distance and near by Worth-4-dot, stereopsis of 140 seconds of arc, and no angle kappa. Intraocular pressure was normal in both eyes. Slit lamp biomicroscopy showed clear corneas, quiet anterior chambers, and bilateral symmetrical dense AIMs, with a very poor pharmacological dilation response in both eyes (Fig. 1a/b). Evaluation of lens clarity, macular appearance, and optic nerve structure were hence very limited. Cycloplegic refraction was −0.75 + 0.75 axis 165° in the right eye and −6.50 + 2.25 axis 165° in the left eye. Retinoscopy was not possible while subjective refinement was inconsistent. Anterior segment ultrasound biomicroscopy (AS-UBM) showed no evidence of iris-lenticular adhesions or lens opacification in either eye (Fig. 2a/b), and B-scan displayed no gross retinal or optic nerve anomalies.

Fig. 1

Anterior segment photography, captured using a Nikon D3300 and Zeiss photo slit lamp: right eye before (a) and after (c) surgery; left eye before (b) and after (d) surgery.

Fig. 2

Ultrasound biomicroscopy of the anterior segment using a Sonomed VuMax before surgery in the right (a) and left (b) eyes, respectively.

The patient and his family were keen to pursue intervention, although caution was advised given the subjective decline in vision of unclear etiology, significant anisometropia, and risk of complications (e.g. hyphema, cataract creation or exacerbation). Despite this, the patient wished to proceed, so bilateral sequential surgical membranectomy was planned, starting with the left eye on one day and proceeding to the right eye at a later date only if symptomatic improvement without complication was achieved in the first eye. Pre-operative and intra-operative sympathomimetic agents (i.e. phenylephrine, epinephrine) were avoided to allow identification and intraoperative treatment of bleeding from severed iris vessels. Viscoelastic was placed under the AIM to stretch the fronds attached to the iris collarette, break any diaphanous connections to the lens capsule that were not detectable on AS-UBM, and create space to protect the lens. AIM fronds were severed at the iris collarette junction using 23G microscissors (MST, Redmond, WA) and the AIM was extracted en-masse through a 1.0mm paracentesis using 23G microforceps (MST). Minimal bleeding occurred in either eye without need for intraocular cautery. BCVA in the patient's left eye at his post-operative week 2 visit improved to 20/30−3 with pinhole, with no evidence of post-operative complications (Fig. 1d). Hence the decision was made to proceed with surgical membranectomy of the right eye 4 weeks later (Fig. 1a/c) using the same approach. Post-operative eye drops in each eye included prednisolone acetate 1% four times daily for 1 week followed by a taper by 1 drop per week, and moxifloxacin 0.5% four times daily for 1 week, but no cycloplegic or miotic agents.

On his last post-operative visit, BCVA improved to 20/20−3 in the right eye and 20/30−2 in the left eye. Final refraction was −2.50 + 0.75 axis 025° and −6.25 + 2.50 axis 175°, respectively. Stereopsis improved to 20 seconds of arc. Both lenses were clear, with normal retina and optic nerve in the right eye, and a mild myopic retinal appearance with tilted disk and scleral crescent in the left eye. His nyctalopia, issues with ambient lighting, visual confusion at near and intermittent torticollis all fully resolved following surgery.

Discussion

Accessory iris membranes are a rare entity that may cause reduced visual function via multiple mechanisms. Our patient presented with somewhat atypical symptoms, including a subjective gradual decline in vision over 2 years that did not improve with refraction, more difficulty seeing at night, and perceiving different images out of his 2 eyes at near, which he could resolve with a mild head turn. All resolved following AIM removal, suggesting a causal association.

Given reduced vision in both eyes, the density and central involvement of his AIMs, as well as his anisometropia, we suspected a component of deprivation and/or refractive amblyopia, which was supported by a review of old notes indicating reduced but relatively stable visual acuity over the past several years. The absence of a subjective or objective decline in acuity following pharmacological dilation or reduced ambient lighting argued against a role for pinhole or stenopaic effects from his AIMs augmenting his vision, while the lack of improvement in acuity following dilation ruled out reduced retinal illumination from his AIMs as an impediment to his vision. While his poor dilation response somewhat clouded this interpretation, we would expect such effects to have remained stable over time, and hence sought at an alternate cause of his subjective decline in vision. Despite our best attempts, we have to accept that suboptimal refraction may have at least partially contributed to his subjective decline in vision, given that both the sphere and cylinder axis changed significantly after surgery in his better-seeing right eye. Given his subjective decline in vision and increasing issues seeing in the dark, cataract development was a concern, although there was no evidence of this in either eye after surgery. While reduced vision and nyctalopia can also result from a retinal dystrophy, there was no sign of this on post-operative examination, and full resolution of his symptoms after surgery would not be expected for this etiology. Hence, we did not perform a work-up for this post-operatively.

Closely-approximated AIM apertures can cause monocular diplopia that varies with working distance owing to the Scheiner principle. Our patient clearly comprehended but denied any double vision, but did endorse seeing different images from each eye at near (visual confusion). This was confirmed by his lack of diplopia at either distance or near on Worth-4-dot. Clinical examination and photographs of his AIMs (Fig. 1a/b) showed no apparent aperture in the pupillary axis of either eye (i.e. the axis connecting the center of the cornea, pupil and fovea), with the only clear opening in each eye being through the temporal pupil (Fig. 1/2). We speculate that his visual axis may have included use of his nasal parafoveal retina in each eye. Given fusion on Worth-4-dot and gross stereopsis pre-operatively, his brain appeared to integrate the visual inputs from each eye (a form of anomalous retinal correspondence using extrafoveal retina in each eye). In such a scenario, one may expect to observe orthophoria on cover testing, fusion on Worth-4-dot and gross stereo, all of which were noted in our patient pre-operatively. One may also expect to observe a negative angle kappa in each eye. This was not identified in our patient, although obscuration of each pupil by his AIMs made identification of a small angle challenging. We speculate that the precise orientation of his eyes during convergence at near, and/or the strain of prolonged near work in college, is what precipitated his visual confusion. His mild left head turn may have facilitated use of his dominant right eye. Furthermore, facultative use of his parafoveal retinas pre-operatively, with conversion to foveal function after AIM removal, may explain his robust acuity and stereopsis gains.

From a surgical perspective, we forewarned the patient about a high likelihood of limited visual improvement in one if not both eyes after surgery as a result of amblyopia. We raised the possibility of a need for cataract surgery if this was the cause of his subjective decline in vision and/or night vision issues. We discussed the chance of unmasking posterior segment anomalies, particularly in a case such as his with unusual presenting symptoms. Despite these concerns, the patient and his family still wished to pursue surgery. Proceeding with sequential surgery with an adequate time interval between surgeries to evaluate for overt symptomatic and/or objective improvement in the first eye provided confidence in the potential yield of surgery for the fellow eye.

Avoidance of pre-operative or intraoperative sympathomimetic agents that may vasoconstrict vessels in the AIM fronds intraoperatively minimized the risk of post-operative bleeding once the constriction wore off. Placement of viscoelastic under the AIM to balloon it away from the lens reduced the risk of accidently touching the lens during Frond lysis or AIM removal, especially if the anterior chamber suddenly shallowed for any reason. Front lysis can be achieved using simple mechanical traction (e.g. with a Sinsky hook). This gives a less-controlled lysis point along the frond, may increase the risk of bleeding if the frond contains vessels, and may unnecessarily stretch the iris at the base of dense fronds. In contrast, use of 20G or 23G microscissors and microforceps allows precise frond lysis with minimal iris manipulation for an optimal cosmetic result and lowers the risk of post-operative inflammation in pediatric eyes. Their use also minimizes corneal incision size, which can be particularly important in young children with more elastic eyes and greater risk of post-operative eye rubbing. Their use only modestly adds to the surgical time, with complete frond lysis and removal achievable in 10–15 minutes using this approach. We also suggest that 23G intraocular cautery is available in the operating room if any significant bleeding develops.

Conclusions

This case highlights that surgical removal of AIMs can provide significant symptomatic, acuity and stereopsis gains, even in the setting of anisometropia, that optimal subjective refraction is challenging even in high-functioning adults with AIMs, and that a subjective decline in vision may not portend neutral or negative outcomes after membranectomy. However, it should be noted that such improvements may not be universal, and each case needs to be examined separately from prior experience and outcomes to avoid unnecessary interventions and/or failed expectations.

Patient consent

Written consent was not obtained as the report contains no personal identifying information.

Funding

No funding or grant support.

Authorship

All authors attest that they meet the current ICMJE criteria for Authorship.

Declaration of competing interest

The following authors have no financial disclosures: OL, KM, RM, JJD.