Table 2.
Evidence tiers for candidate prognostic factors in POMS, with the outcome each factor predicts
| Factor | Outcome predicted | Tier | Rationale/caveat |
|---|---|---|---|
| Relapse number (first 2 years); short inter-attack interval; ARR | Time to EDSS 3.0/4.0/6.0; SPMS conversion | A—Replicated | Consistent across ≥ 4 independent cohorts, multivariable-adjusted [19–21] |
| EDSS change in first 2 years | 9-year EDSS; disability worsening | A—Replicated | Large effect sizes; distinct from baseline EDSS [22] |
| Brainstem and spinal cord lesion topography; ≥ 2 new T2 lesions | 9-year EDSS; relapse rate | A—Replicated | Concordant with adult MS; single large POMS cohort plus CIS data [22, 23] |
| Serum NfL | MRI/clinical activity; treatment escalation | A—Replicated | Multiple designs incl. RCT post hoc [24–27]; needs paediatric age-adjusted Z-scores and assay harmonisation |
| DMT delay; early high-efficacy therapy | Sustained EDSS 4.0; time to first relapse; disability worsening | A—Replicated | Registry-based [28–30]; subject to confounding by indication |
| Thalamic volume; NAWM integrity; infratentorial lesion number | Time to first relapse; 6-month CDW | B—Preliminary | Single 12-year cohort, n = 52 [31]; needs paediatric normative data and external validation |
| Serum GFAP; CHI3L1; CSF kappa free light chains | Relapse activity; progression | B—Preliminary | Strong adult rationale [32–35]; minimal paediatric data [36] |
| Cognitive impairment at onset; cognitive reserve | Cognitive trajectory; socio-professional attainment | B—Preliminary | Single 12-year cohort, n = 48 [10]; reserve rarely measured [37] |
| Sleep quality; psychiatric comorbidity; social determinants | Fatigue; cognition; MRI outcomes | B—Preliminary | Cross-sectional or associative [38–42] |
| Dietary fat/vegetable intake; physical activity; vitamin D status after diagnosis | Relapse rate | B—Preliminary | Post-diagnosis outcomes but self-reported exposure, single cohorts [43–46] |
| BMI/obesity; breastfeeding; caesarean delivery; maternal illness; parental occupational exposure; air pollution; household pesticides; EBV/HSV-1; passive smoke | Risk of developing POMS; age at onset | C—Susceptibility | Not shown to predict post-diagnosis course [47–56]; relevant to prevention, not prognostic modelling |
| Baseline EDSS | Long-term EDSS vs confirmed improvement | D—Inconsistent | Opposite directions across cohorts [22, 28]; plausibly regression to the mean and unstandardised assessment timing |
| Gadolinium-enhancing brainstem/cervical lesions at year 1 | 9-year EDSS | D—Inconsistent | Negative coefficient unexplained [22]; likely detection timing and treatment response rather than protective inflammation |
| Absence of mental-state involvement at onset | Third attack; EDSS > 4 | D—Inconsistent | Plausibly diagnostic misclassification (ADEM) in older cohorts [20, 16, 57] |
ADEM acute disseminated encephalomyelitis, ARR annualised relapse rate, CDW confirmed disability worsening, CHI3L1 chitinase-3-like protein 1, CIS clinically isolated syndrome, DMT disease-modifying therapy, EDSS Expanded Disability Status Scale, GFAP glial fibrillary acidic protein, NAWM normal-appearing white matter, NfL neurofilament light chain, POMS paediatric-onset multiple sclerosis, SPMS secondary progressive multiple sclerosis