Optic atrophy secondary to minocycline-induced idiopathic intracranial hypertension

Discussion

CARP (Gougerot-Carteaud syndrome) is a rare skin disorder characterised by hyperpigmented macules and patches that coalesce into reticulated (net-like) patterns and papillomatous appearances. It often appears on the chest and upper back and appears similar to tinea versicolor. However, fungal staining of skin scrapings is negative, and it has no response to antifungal treatment. The incidence and prevalence of CARP are not known in the USA, but retrospective reviews have identified its mean age of onset as 15 years old, with most cases diagnosed between 8 years and and 32 years of age.

First-line therapy for CARP is minocycline (typically 50–100 mg two times per day for 4–6 weeks) due to its excellent response. Second-line therapy consists of macrolide or alternative antibiotics—azithromycin, erythromycin, clarithromycin, roxithromycin, tetracyclines and cefdinir—for patients who cannot tolerate, fail to respond or are contraindicated (eg, young children) to minocycline. Alternative therapies for refractory cases include topical retinoids, tacrolimus, and vitamin D analogues.14 15

The mechanism of how minocycline causes increased ICP is likely related to disruption of secondary cell signalling—cyclic adenosine monophosphate—in the arachnoid villi, which are responsible for CSF absorption. Additionally, minocycline is highly lipophilic, allowing it to cross the blood–brain barrier more effectively.16 Decreased CSF absorption can lead to increased ICP that, when compounded, can result in IIH.

Minocycline-induced IIH has been reported in the literature many times. It has a half-life of 24 hours and symptoms related to IIH are often reported during use.6–13 However, prior minocycline use can present with elevated ICP up to 5 weeks after cessation.17 Despite the elevated opening pressure on LP, there was no significant presentation of headache prior to admission. Visual and hearing symptoms did not present until 4–5 weeks after the last minocycline dose.

This demonstrates the need for close monitoring for papilloedema during minocycline use. Our patient did not have a baseline ophthalmic exam prior to minocycline use. In addition, no ophthalmic exam was recommended during use or after her minocycline prescription. ICP likely increased significantly in the interim and eventually led to papilloedema, but due to lack of ophthalmic monitoring, it was not noticed in time. This atypical course of minocycline-induced IIH in terms of symptoms and timeline warrants the need for closer clinical monitoring for papilloedema secondary to IIH to help prevent the most serious complication of IIH, which is consecutive optic atrophy with irreversible visual loss and very poor prognosis as in this case. Timely monitoring and ophthalmic examination could have helped prevent this pitfall.

In terms of management, the first step is always a diagnostic LP to confirm the opening pressure. This can often be therapeutic. The mainstay of medical management is acetazolamide or topiramate and cessation of the offending medication (eg, minocycline) if possible. Repeat LPs are often used as therapy. Weight loss is helpful but can be a slow process, hence not helpful in acute situations.

Acetazolamide, a carbonic anhydrase inhibitor, is usually the first-line medical therapy followed by topiramate, although it is less efficacious. If patients are unable to tolerate it, furosemide, another diuretic is often used.18

Surgical intervention is usually considered if medical treatment fails or if the initial visual loss is extremely severe, as in our case. Surgical intervention for IIH remains controversial in terms of preferred methods. The two main options are CSF shunting or optic nerve sheath fenestration (ONSF). Retrospective case series of either procedure for IIH have demonstrated similar outcomes.19 However, there are no current randomised controlled trials demonstrating a more effective surgical treatment based on predominant symptoms.20 Another option, and an increasingly recognised intervention, is dural venous sinus stenting (DVSS). This third option targets one of the common abnormalities associated with IIH: venous stenosis. Regardless of choice, all procedures have variable efficacies and associated adverse events.

ONSF is usually the preferred method for predominant visual symptomatic relief but minimal headache,21 which would have been ideal in this case, but given evidence of such high ICP, our patient underwent a shunt placement instead.

The main goal of ONSF is preservation of vision with complications ranging from transient diplopia to permanent vision loss in 2.6% of cases.22 Previous retrospective reviews have shown improvement in headaches (44%), vision (59%) and papilloedema (80%) with ONSF.23 In acute situations associated with significant visual loss without much headache, like our case, ONSFs are recommended. The complications/risks or side effects of ONSF in the paediatric population have not been extensively studied. However, previous case reviews have reported uncomplicated outcomes.24 25 While no specific criteria differentiates complicated versus uncomplicated outcomes, uncomplicated or minor complicated outcomes have generally been described in adults as diplopia, late failure, corneal dellen, synechiae, atonic pupil, tonic pupil, anisocoria, pupillary dysfunction, perilimbal conjunctival bleb, peripapillary haemorrhages, disc haemorrhage, cyst formation and conjunctival abscess.23

Our patient underwent a CSF shunting procedure a few days after her LP due to lack of symptom relief. The most common indication for CSF shunting procedures is severe headache or progressive visual loss. In most cases, shunting relieves most IIH symptoms, including vision restoration. For example, previous meta-analyses, including CSF shunting, have shown improvement in headaches (80%), vision (54%) and papilloedema (70%) post procedure. However, the main complication of lumboperitoneal shunt is shunt failure, occurring in about 43%–50% of patients.23 26–29 Shunt failure in rare cases can lead to worsening ICP with worsening vision secondary to optic nerve ischaemia.30 In this case, the patient developed shunt failure, leading to further optic nerve damage, which caused further optic atrophy. The shunt was revised with continuation of medical therapy before she finally saw neuro-ophthalmology. Of note, it would have been an option to reset her shunt settings instead of restarting her on acetazolamide again.

Finally, DVSS is an increasingly recognisable option. Although only indicated for a small percentage of patients with IIH, recent retrospective comprehensive reviews of DVSS have shown it to be a very successful procedure, close to 100%, with relatively low rates of repeat procedures, 10%–18%.23 28 29 Additionally, the procedure is minimally invasive compared with CSF shunting or ONSF. The main indication for it is based on radiologic evidence of dural venous stenosis and an increased pressure gradient across the stenosis, often >8 mm Hg. In terms of clinical outcomes, it appears DVSS has significant improvement in both visual (78%) and headache symptoms (82%–83%).23 28 29 Of note, bilateral transverse sinus stenosis is present in 90% or more of patients with IIH which makes DVSS an attractive and potentially more effective option for the subset of IIH patients who meet the interventional criteria.

In short, despite the appropriate indicated interventions for fulminant IIH and papilloedema, our patient did not recover as expected. Fraser et al reported a similar case with worsening visual loss and fulminant papilloedema even after stopping minocycline, which was finally treated with surgical intervention. A study by Winn et al highlighted that despite a half-life of less than 24 hours, minocycline can keep the ICP elevated for at least 2–5 weeks.7 17 Our case reports an even longer duration of increased ICP after the last dose of minocycline, which required her to remain on acetazolamide and lumboperitoneal shunt placement even though there was documented weight loss. She unfortunately had shunt failure, which led to a confounding picture of worsening ICP and visual symptoms. Killer et al has also reported a similar case where there was reported persistent papilloedema despite VP shunt placement—postulating a lack of bidirectional flow of CSF at the optic nerve or optic nerve compartmentalization as the underlying culprit—although there was no shunt failure in their case.31 Our patient did, however, have initial subjective visual recovery in the first few months after the lumboperitoneal shunt but, unfortunately, had shunt complications.

In conclusion, due to widespread prescription and use of minocycline for common dermatologic conditions, the severe risk of IIH should be emphasised to all patients, even though it is rare, since visual loss is irreversible. Care should be coordinated with ophthalmologists and primary care physicians to monitor the development and possible prevention of permanent vision loss.