TP53 variant function and allele fraction split monoallelic MDS into distinct risk groups

Design

Retrospective analysis of two independent MDS cohorts (IWG and J-MDS)

Population

4,505 patients with MDS; 271 monoallelic and 499 multi-hit TP53

Primary outcome

Leukaemia-free survival by TP53 allelic state and variant function

Effect

High score + VAF 22% or more: 9.2 months; low score + VAF under 22%: 44.7 months

A two-cohort analysis in Blood (1 October 2026) studied pretreatment samples from 4,505 patients with myelodysplastic neoplasms, including 271 with a monoallelic TP53 mutation and 499 with multi-hit TP53. Each variant was annotated with a published functional phenotype score from saturation mutagenesis, reflecting dominant-negative and loss-of-function effects.

Median leukaemia-free survival was 37.2 months with wild-type TP53, 19.2 months with monoallelic and 7.8 months with multi-hit disease. Within the monoallelic group, a high phenotype score (7 or more) meant much shorter survival than a low score (12.6 vs 43.3 months), particularly in patients classed as lower risk by IPSS-R and IPSS-M. Combining score and allele fraction sharpened this: high score with VAF 22% or more behaved like multi-hit disease (9.2 vs 7.8 months), while low score with VAF under 22% behaved like wild-type (44.7 vs 37.2 months).

For the laboratory, this means the variant itself, its allele fraction and an assessment of allelic state all carry prognostic weight, and a report that says only 'TP53 mutated' leaves out what the haematologist needs. The phenotype score is not yet part of IPSS-M or WHO criteria, so it should inform discussion rather than reclassify patients.

• Report the exact TP53 variant, its allele fraction and any evidence of multi-hit status
• Check for TP53 deletion or copy-neutral loss of heterozygosity to establish allelic state
• Flag monoallelic cases in lower-risk MDS for discussion, where variant function changed outcome most
• Keep IPSS-M and WHO criteria as the formal risk framework; the phenotype score is not yet in them

Why it matters

It challenges treating every monoallelic TP53 mutation as a single intermediate-risk category.

Don't overread it

The phenotype score and VAF cut-off are not yet part of formal risk scores and need external validation.

The statistics, in plain English

The p-values of 0.3 and 0.8 for the combined groups mean those subsets were statistically indistinguishable from multi-hit and wild-type disease respectively, which is the point. Subgroups within 271 monoallelic patients are small, so cut-offs such as VAF 22% need confirming in other cohorts.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.