Untreated LHON Vision Loss Proves More Dynamic Than Previously Thought

A large retrospective study of people with Leber hereditary optic neuropathy (LHON) who never received the approved drug idebenone has mapped, in unusual detail, how their vision changed over time without treatment. The CRS-2 study, published in the Journal of Neurology, analyzed case records from 219 patients across 20 centers in seven European countries and found that spontaneous visual acuity changes—including meaningful recovery—occurred not only early in the disease but well into its so-called chronic phase. The findings challenge the classic picture of LHON as a condition in which vision collapses rapidly, stabilizes at a poor level within about a year, and then stays fixed.

LHON is an inherited mitochondrial disorder that causes subacute, bilateral, painless vision loss, typically beginning in one eye with the second eye becoming involved within weeks or months. More than 90 percent of cases stem from one of three mitochondrial DNA variants: m.3460G>A in MT-ND1, m.11778G>A in MT-ND4, and m.14484T>C in MT-ND6. These variants impair complex I subunits of the mitochondrial respiratory chain, creating an energy deficit that hampers axonal transport, while accumulated reactive oxygen species may drive retinal ganglion cell death. Symptom onset usually occurs between 15 and 35 years of age, with a strong male predominance, and prevalence estimates from several countries range from roughly 1 in 31,000 to 1 in 68,000.

The CRS-2 study was designed to characterize the natural history of vision loss and spontaneous recovery in genetically confirmed LHON, and it also served as a repository of treatment-naïve control data for the LEROS trial of idebenone. Researchers reviewed case records from 1999 to 2018, collecting data between May 2016 and March 2018. To qualify, patients had to be at least 12 years old, carry one of the three common pathogenic variants, have symptom onset after 1999 with the month known for both eyes, and have at least two dated visual acuity assessments within five years of onset before any idebenone use. Patients in interventional trials, or with other causes of visual impairment or active eye disorders, were excluded.

The cohort was demographically typical of LHON: 79.9 percent of the 219 patients were male, a 4:1 male-to-female ratio, and just over half had first symptom onset between 15 and 35 years of age, with a median age of 26.0 years. The m.11778G>A variant accounted for 71.7 percent of patients, m.14484T>C for 14.6 percent, and m.3460G>A for 13.7 percent. Symptoms developed in both eyes simultaneously in 45.7 percent of patients; among those with sequential onset, the median delay between eyes was two months. Idebenone use was recorded for 113 patients after their baseline visit, and data from that point onward were excluded to keep the analysis genuinely treatment-naïve.

After exclusions, 402 eyes were available for analysis, of which 385 were in the subacute or dynamic phase—within one year of symptom onset—at baseline. The researchers defined four spontaneous outcomes aligned with the LEROS study: clinically relevant recovery (improvement from off-chart vision to reading on-chart, or a gain of at least 10 letters for those already on-chart), clinically relevant stabilization (maintaining visual acuity better than 1.0 logMAR), clinically relevant benefit (a composite of recovery or stabilization), and clinically relevant worsening (dropping off-chart or losing at least 10 letters). Off-chart values were standardized to 1.8 logMAR to remain consistent with LEROS methods.

The primary endpoint—the proportion of subacute/dynamic eyes with clinically relevant benefit 12 months after baseline—was met by 20.8 percent of eyes (20 of 96), driven by 12 eyes achieving recovery and 11 achieving stabilization. In the smaller 24-month analysis of 44 eyes, 43.2 percent showed clinically relevant benefit. Outcomes varied sharply by mitochondrial variant. Eyes with m.14484T>C fared best, with recovery rates of 50.0 percent at 12 months and 66.7 percent at 24 months, though numbers were small. In the 12-month analysis, m.3460G>A eyes showed the worst results, with no recoveries and worsening in 81.8 percent, whereas at 24 months the m.11778G>A variant carried the poorest outlook.

Age and sex also mattered. Eyes of patients younger than 15 at symptom onset had the best outcomes, with the highest recovery rates and the lowest worsening rates in both analyses, and regression analysis confirmed their visual acuity change was significantly more favorable than the 15–35 years reference group. Eyes from female patients showed slightly worse outcomes than those from males, a difference that reached statistical significance only in the 24-month regression. Notably, 67 eyes achieved spontaneous recovery at their last assessment at a median of 3.4 months from baseline, but a quarter of them did so 12.8 months or later—evidence that meaningful improvement can arrive long after the early disease phase.

Perhaps the most striking findings came from visual acuity trajectories. Among 32 eyes followed for more than a decade, the nadir of vision did not always occur within the first year, and acuity changes continued well into the chronic stage, with substantial variability both between and within patients. Among 37 eyes assessed at both 12 and 24 months, nearly a quarter changed blindness category between the two timepoints, with the overall proportion declining in the better-vision category and increasing in the off-chart category. The authors caution, however, that the 12- and 24-month analysis populations differed, so their rates should not be read as longitudinal changes in the same eyes.

The study has important limitations inherent to its retrospective design. Follow-up regularity and duration varied widely, ranging from 0.1 to 196.4 months, and factors such as smoking, alcohol consumption, supplementation, and stress were not recorded. Testing procedures, chart types, and distances were not prospectively harmonized across sites, and off-chart improvements or deteriorations could not be quantified in logMAR terms. Changes in referral pathways and diagnostic awareness over the long inclusion period may have introduced ascertainment bias. A post hoc analysis using a stricter 15-letter threshold, however, barely changed the outcome rates, supporting the robustness of the main results.

By comparing CRS-2 with the LEROS natural history group, the researchers found closely matched outcomes overall—20.8 percent versus 20.7 percent achieving clinically relevant benefit at 12 months—with the notable exception of m.3460G>A eyes, which fared worse in CRS-2 and may have been underrepresented among matched controls. The team concludes that CRS-2 offers the largest contemporary dataset on untreated LHON visual acuity and suggests the disease course is more dynamic than previously reported. These findings provide a benchmark against which therapeutic outcomes can be interpreted, underscore the need for caution when relying on nadir or last-visit values, and highlight the importance of timing when evaluating spontaneous recovery in this rare optic neuropathy.

Subject of Research: Natural history of visual acuity changes in untreated Leber hereditary optic neuropathy

Article Title: Visual acuity changes in idebenone treatment‑naïve patients with Leber hereditary optic neuropathy: insights from the CRS-2 study

Article References: Krawczynski, M. R., van Everdingen, J., Leroy, B. P., Verny, C., Bauer, N. J. C., Andris, C., Lorenz, B., Votruba, M., Priglinger, C., Jurkute, N., Llòria, X., Silva, M., Carelli, V., Yu-Wai-Man, P., & Klopstock, T. (2026). Visual acuity changes in idebenone treatment‑naïve patients with Leber hereditary optic neuropathy: insights from the CRS-2 study. Journal of Neurology, 273(11), Article 690. https://doi.org/10.1007/s00415-026-14181-6

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14181-6

Keywords: Leber hereditary optic neuropathy, CRS-2 study, visual acuity, spontaneous recovery, mitochondrial DNA variants, idebenone, natural history, retrospective study, m.11778G>A, m.14484T>C, logMAR, chronic phase