Optic Neuritis and the Cost of Faster Vision Recovery
A single patient with acute optic neuritis and a clean brain MRI. High-dose IV corticosteroids won’t change her eventual vision — ONTT already established that — but they will get her back to functioning faster. The debate is whether that speed is worth a real glycemic cost for a guardian who can’t afford to be sidelined.
R.S., a 34-year-old woman, teaches second grade at the same elementary school she attended as a child, and has been the sole guardian for her younger sister — sixteen, cerebral palsy, non-ambulatory — since their mother’s death three years ago. Three days ago she noticed her right eye ached with any movement, worse looking up or to the side, and the ache was quickly followed by blurring she first blamed on end-of-day fatigue grading spelling tests. It hasn’t cleared. This morning she could no longer read the whiteboard from the front row of her own classroom, and colors on her desk — red folders especially — looked washed toward gray.
She has had type 1 diabetes for ten years, managed on an insulin pump, with an HbA1c around 7.1% that her endocrinologist has called genuinely good control — no retinopathy, no nephropathy, no prior hospitalization for either hypo- or hyperglycemia. Nothing else in her history is remarkable. On exam her visual acuity in the right eye has fallen to 20/100 from a baseline near 20/20, she has a clear relative afferent pupillary defect, and red objects held before that eye look distinctly duller than the same object held before the left. A brain MRI obtained this morning, ordered as part of the standard optic neuritis workup, shows no periventricular or juxtacortical white-matter lesions — a genuinely clean scan, not an equivocal one.
That clean MRI matters more than a reassuring aside. The Optic Neuritis Treatment Trial established two things that pull in different directions for a patient like her: IV methylprednisolone speeds visual recovery measurably in the first weeks, but by six months, visual outcomes converge regardless of whether she is treated at all — the steroid buys time, not final acuity. The trial’s long-term follow-up also found a real, if fading, reduction in two-year MS conversion risk from IV treatment, but only in patients whose baseline MRI already showed white-matter lesions. Her scan has none. Whatever argument exists for treating her, the MS-risk argument is not it — she sits outside the population where that particular finding ever applied.
So the actual question in front of the team is narrower than whether optic neuritis gets IV steroids — ONTT already answered that generally. It is whether accelerating a recovery that is coming to the same place either way is worth several days of high-dose glucocorticoid on top of an insulin pump she has kept in careful balance for a decade, for a woman whose sister needs her functioning now, not just eventually recovered.
In clinic, the morning her vision failed to improve
I want to start her on IV methylprednisolone today, and I want to be honest about exactly what it will and won’t do. ONTT is unambiguous that six-month visual outcomes look the same whether or not she’s treated — I’m not going to overstate that. What it also showed is a real, measurable difference in how fast function comes back in the first two to three weeks, and for R.S. specifically, those two to three weeks aren’t abstract. She’s the only adult in her house. A slower recovery isn’t a cosmetic loss for her — it’s weeks of being functionally unable to read to a sister who depends on her for basically everything.
You’re right that the weeks matter more for her than for most patients — I’m not going to argue her circumstances aren’t real.
But “ONTT showed faster recovery” is doing a lot of work in that argument, and it’s worth being precise about what it doesn’t show: it doesn’t show any long-term benefit for a patient whose MRI is clean, which is exactly her situation. What we’re actually weighing isn’t steroids versus nothing — it’s three days of high-dose glucocorticoid, on top of an insulin regimen she’s kept genuinely stable for a decade, purely to buy speed with no outcome difference behind it. I’ve watched pump-dependent patients swing badly on a pulse like this, and the corrective work afterward isn’t trivial either.
I don’t think this actually has to be a binary choice between the recovery-speed benefit and her glucose control, because the second problem is solvable in a way the first one isn’t. If we bring her endocrinology team in before the first dose — not after a glucose reading surprises everyone — we can pre-adjust her basal rates and build a correction protocol into the same three days. That doesn’t erase the risk, but it converts it from an argument against treating her into a logistics problem we can actually manage. And I’d rule out oral prednisone alone as any kind of compromise here — ONTT found that arm carried a higher rate of recurrent optic neuritis than either IV steroids or no treatment at all, so a gentler version of the same drug isn’t actually gentler, it’s a worse choice on a different axis entirely.
Agreed within the visit: IV methylprednisolone today, per the ONTT protocol, with endocrinology looped in before the first dose to pre-adjust her basal insulin rates and set a same-day correction protocol rather than reacting to numbers after the fact. Oral prednisone taper to follow the IV course; oral steroid alone ruled out entirely, not offered as a gentler alternative.
Not agreed, and not really expected to be: whether the whole intervention was worth doing in the first place, given that her six-month vision will look the same either way. The neuro-ophthalmologist and the endocrinologist left the visit holding genuinely different weightings of the same facts — one reading “no outcome difference” as reason enough not to accept any glycemic risk, the other reading “weeks of functional vision, for this patient, is the outcome that matters” as reason enough to accept a manageable one. Neither treated the other’s position as a misreading of the evidence, only as a different judgment about what the evidence is worth to her specifically.