- Design
- Cross-platform diagnostic comparison
- Population
- 23 paediatric tumours (18 CNS, 5 non-CNS)
- Primary outcome
- Agreement of methylation class with integrated histology
- Effect
- Family-level agreement in all included CNS cases; MGMT status concordant 16/17; Rapid-CNS2 class agreement 16/17
A study in Modern Pathology (7 September 2026) profiled 23 paediatric tumours, 18 of them CNS, on both Illumina EPIC methylation arrays and Oxford Nanopore sequencing, and compared classifier outputs with integrated histology.
In CNS tumours, the two platforms correlated closely apart from one outlier, which was excluded. Classification at family level agreed with histology in every remaining case, copy-number profiles were concordant, and MGMT promoter status matched in 16 of 17. Among nanopore pipelines, Rapid-CNS2 gave the best class-level agreement, 16 of 17. In five non-CNS tumours, the pan-cancer classifier performed poorly, agreeing in only one.
The study is small and excluded an outlier, and nanopore needs fresh-frozen tissue. Its point is speed: a methylation class the same day rather than weeks later from a reference laboratory. For centres in India without array access, a nanopore pathway could shorten time to diagnosis, if fresh tissue handling can be arranged.
- Treat nanopore methylation as reliable for family-level CNS tumour class in this study, not yet for every class.
- Plan fresh-frozen tissue collection at surgery if nanopore profiling is to be used.
- Do not rely on the current pan-cancer classifier for non-CNS tumours.
- Integrate any methylation result with histology and other molecular findings before reporting.
Why it matters
Turnaround, not accuracy, is what keeps methylation classification out of reach for many centres.
Don't overread it
Twenty-three tumours, one excluded outlier and no non-CNS validation; this is not yet a replacement for arrays.
The statistics, in plain English
Agreement of 16 of 17 sounds high, but in so few cases a single disagreement moves the figure by six points, and the confidence interval would be wide.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for pathology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free