Optic nerve assessments in MS diagnosis are worth the added implementation complexity: No

The diagnosis of multiple sclerosis (MS) has historically relied on clinical grounds. This involved identifying at least two attacks with either clinical or historical evidence in different anatomical locations, provided there was no better explanation for the symptoms. ¹

In this context, optic nerve (ON) involvement could be confirmed through common neurological signs. These include a relative afferent pupillary defect, indicating demyelination of the affected ON compared with the unaffected eye, or a pale optic disc observed during fundoscopy, consistent with ON atrophy.

As medical science advances, more precise biomarkers are crucial for confirming semiological findings and aiding differential diagnoses. This is especially true for disorders lacking a single sensitive or specific test, where consensus criteria are vital for early diagnosis, treatment, and reducing detrimental misdiagnoses.

Objective evidence of both acute and chronic demyelination can enhance the diagnostic accuracy of MS. This evidence includes prolonged visual evoked potentials (VEP), thinning of the peripapillary retinal nerve fibre layer (pRNFL) or ganglion cell and inner plexiform layer (CGPL) as detected by optic coherence tomography (OCT), and presence of demyelinating lesions in ON-Magnetic Resonance Imaging (MRI) with fat saturation. Nonetheless, the need for qualified personnel, established methodologies, and strict quality control measures to accurately interpret the results might not justify the implementation complexity of these techniques from a global standpoint.

Orbital MRI is useful in selected clinical scenarios, ² but its applicability is limited because it requires dedicated high-resolution sequences (increasing acquisition time by about 20 minutes per patient, although shorter protocols have also been proposed to be useful in specialised centres), access to appropriately equipped scanners, and interpretation by neuroradiologists with specific expertise.^2,3 Moreover, ON-MRI may fail to detect abnormalities in approximately 5% to 25% of patients with clinically typical MS-associated optic neuritis, reducing its diagnostic sensitivity and challenging its routine implementation. ³ Variability in sensitivity and the risk of misclassification in non-specialised settings further argue against its systematic use.^2,3

Similarly, although OCT and VEP can detect and quantify neuroaxonal damage, their diagnostic reliability depends heavily on strict acquisition and interpretation conditions. ⁴ Accurate OCT assessment requires adherence to OSCAR-IB quality control criteria, use of consistent device platforms, and access to neuro-ophthalmology expertise, ⁴ none of which are uniformly available in routine clinical care. ⁵ Furthermore, its diagnostic usefulness is most apparent when conducted more than 3–6 months after an acute event. This delay neutralises the potential advantage of earlier diagnosis offered by the 2024 McDonald Criteria.^4,6

VEP can demonstrate delayed conduction supporting evidence of optic pathway involvement; however, the exact measures depend on technical and methodological factors, and are also centre-dependent, which further limits reproducibility across clinical environments. ⁴ In addition, common ophthalmologic comorbidities, such as glaucoma, myopia, and ischemic or compressive neuropathies, may mimic or obscure MS-related damage, increasing the risk of diagnostic misinterpretation. ⁴

The updated 2024 McDonald Criteria represent progress towards a biological diagnosis and earlier treatment for MS, reaching patients previously identified with clinically or radiologically isolated syndrome. However, these new criteria also present new challenges in many regions. Global obstacles, such as limited access to neurologists and paraclinical tools, are already contributing to diagnostic delays and misdiagnosis around the world. ⁷ Updated figures from the 2025 ECTRIMS meeting showed that MRI was available and routinely used in 99% to 98% of the surveyed countries, while OCT varied from 89% (availability) to 27% (routinely used), and VEP was available in 85%, but routinely used only in 52%. ⁸ Our experience in Latin America, where both private and public health systems coexist, echoes this complexity. While specialised centres in both systems report 100% MRI availability, access to neuroradiologists with MS expertise significantly drops to 40% in public centres and 67% in private centres. ⁹ Furthermore, OCT availability is lower still, present in only 80% of private and 44% of public centres. Crucially, among those with OCT, only 50% adhere to the OSCAR-IB criteria. ⁵ This data underscores that not just access, but also stringent quality control, is essential for reliably incorporating these paraclinical tools into new diagnostic criteria. In addition, research is scarce on the cost-effectiveness or cost-benefit of implementing these new diagnostic techniques, particularly in resource-limited settings.

MRI remains the cornerstone of MS diagnosis and long-term prognosis, with stronger evidence for clinical validity, availability, and reproducibility. ² The modified dissemination in space criteria have higher sensitivity (92.5% (with ON-MRI) vs 88.2%), but slightly lower specificity (80.0% (with GCIPL IEA ⩾ 4 μm) vs 82.2%), with overall similar accuracy (86.6% (with ON-MRI) vs 86.5%) than the 2017 dissemination in space criteria. ¹⁰ These modest diagnostic gains from ON assessments, particularly in asymptomatic patients, have yet to be shown to improve outcomes to justify assets reallocation to be included during the routine diagnostic and monitoring processes in resource-limited settings.

Therefore, although OCT and VEP provide complementary value, their generalised, routine implementation in MS diagnosis is not widely justified, and their use should remain context-dependent, reserved for highly standardised settings with the appropriate infrastructure, expertise, and quality assurance measures.

Before we broaden diagnostic protocols to include OCT or VEP in regions where these techniques are not currently available, our focus should be on guaranteeing widespread access to high-quality MRI and ongoing medical education for neurologists specialising in MS diagnosis. This priority stands until formal economic modelling clearly demonstrates the clinical and cost benefits of such expansions.