TP53 mutation analysis in myelodysplastic syndromes - long or short read sequencing?

Diagnosis, prognostication and management of myelodysplastic neoplasms (MDS), like hematology in general, has undergone a paradigm shift over the last few decades. Morphology and blast counting are still relevant, but molecular abnormalities are becoming more relevant.¹ Genetic testing for common myeloid mutations has become an integral part of routine MDS evaluation.^1,2 Practically most laboratories apply short-read sequencing (SRS), as performed by next-generation sequencing (NGS), in which multiple short (50-300 base pairs) DNA fragments are generated and simultaneously read.³ Over the years, we have learnt that it is not only the mutated gene that matters but that other features such as, specific hot spots, the variant allele frequency (VAF), the number of mutations and gene-gene interactions can influence prognosis and treatment.

For years, mutations of the TP53 tumour suppressor gene, the “guardian of the genome”, have been known to be associated with poor outcomes in MDS.^1,4 The landmark study of Bernard et al. has refined this “all-or-nothing” view.⁵ They demonstrated that the adverse prognosis of TP53 is almost exclusively tied to “multi-hit” (biallelic) inactivation (67% of the TP53-mutated cases). Patients with a single TP53 mutation (monoallelic, 33% of MDS patients with TP53 mutations) often exhibit outcomes similar to those with wild-type TP53. In this regard, recent studies have identified mutations in the TP53 regulator PPM1D to be common among patients with therapy-related MDS more than in primary MDS. Furthermore, PPM1D carriers had worse prognosis after stem cell transplantation, and they tended to co-occur with TP53 mutations.⁶ Monoallelic TP53 mutations are defined as a single mutation without concurrent deletion or copy neutral loss of heterozygosity (cnLOH). One could argue that biologically a biallelic mutation in TP53 makes no sense to cells as the mutations have a dominant-negative (DN) effect. However, the selection pressure for biallelic loss, despite the initial DN effect of a single mutation, is a hallmark of clonal evolution. The DN effect, where a mutant TP53 protein “poisons” the P53 tetramer, provides an initial advantage by leaving only a small fraction (statistically 1/16 in a heterozygous cell) of the functional protein. To achieve total “proliferative freedom,” cancer cells benefit from eliminating that last vestige of wild-type activity, which often occurs through LOH or deletion of the 17p arm, resulting in full biallelic inactivation. Bernard’s discovery prompted the recognition of “MDS with biallelic TP53 inactivation” as a distinct, high-risk clinical entity in the new 2022 MDS classifications.^7,8

In practice, determining whether two mutations are on the same chromosome (cis), or different chromosomes (trans), a process called phasing, is important for genetic-clinical decision making. Bernard’s paper raised a problem: SRS can identify two or more TP53 hits, but only if they are close to each other. If two mutations are far from each other, routine NGS cannot tell if they are on the same allele (cis) and mono-allelic or different alleles (trans), biallelic (Table 1). Thus, accurately defining “biallelic” in the clinic remains a challenge. In practice, when an MDS patient is found to have ≥2 TP53 mutations, it raises the question: are these mutations located on the same allele (monoallelic) or on different alleles, indicative of multi-hit TP53 inactivation? This has prognostic relevance, dictating the management. Currently, clinicians rely on surrogates and consider as biallelic any case with multiple mutations, a single mutation with concurrent 17p deletion (via fluorescence in situ hybridization [FISH]), or a high VAF (VAF >50%). While these surrogates are highly predictive, they are not direct evidence of phasing, leaving the question of mono or biallelic status open.

In this issue of Haematologica, Zeuthen and colleagues address this question.⁹ They utilized the Oxford Nanopore platform, and long-read sequencing (LRS) of the TP53 coding regions (exons 2-11) in 29 MDS patients with ≥2 TP53 mutations, 62 mutations totally. They obtained perfect concordance between SRS and LRS for mutation detection (62/62 mutations). Moreover, they could determine the allelic status, and in 28 of 29 patients (96.6%), multiple mutations indeed represented a multi-hit configuration. Only a single patient (3.4%) was found to have two mutations in cis (multi-hit but monoallelic status). Their findings provide a reassuring “reality check” for current diagnostic practice: once SRS demonstrates ≥2 TP53 mutations, they are very likely multi-hit, and you don’t need fancy expensive tools to count or locate those hits. Thus, in a way, for ≥2 TP53 mutations in NGS, “we have come back to square one”, or in other words, indeed, we are dealing with a poor prognostic disease.

Table 1.Allelic state assessment: standard short-reading sequencing versus long-reading sequencing.

LRS also enables detection mutations that cannot be found by SRS and might shed light on the disease biology.¹⁰ However, the main contribution of this paper is determining the allele status of TP53 multiple mutations, and delivering the message that continuing with SRS in routine practice is enough in most cases.

The study has some limitations. The relatively small sample size (N=29), the lack of cytogenetic data for some patients (e.g., del(17p) or copy-neutral loss of heterozygosity) remains a hurdle. Without chromosomal data, a “monoallelic” sequencing result cannot definitively rule out biallelic inactivation occurring at the structural level. Also, one unresolved nuance discussed by the authors is “sub-clonal mosaicism”. There are some technical VAF discrepancies, and lack of clinical correlation. Nevertheless, the pragmatic conclusion of the authors remains: while LRS is technically superior for direct phasing, the high prevalence of true biallelic status in patients with multiple mutations suggests that routine SRS, combined with VAF analysis and cytogenetics, remains sufficient and cost-effective for most MDS patients with multiple TP53 mutations. LRS may be reserved for “selected patients”, those where the VAF is ambiguous or where treatment decisions (transplant ? intensive chemotherapy ?) hinge on the precise definition of the TP53 state.

Footnotes

• Received March 27, 2026
• Accepted April 1, 2026

Correspondence

M. Mittelman moshemt@gmail.com

Disclosures

No conflicts of interest to disclose.

Contributions

Both authors contributed equally.