TP53 ‐Mutated Myeloid Neoplasms: 2024 Update on Diagnosis, Risk‐Stratification, and Management

ABSTRACT

Alterations in the tumor suppressor gene TP53 are common in human cancers and are associated with an aggressive nature. Approximately 8%–12% of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) harbor TP53 mutations (TP53 ^mut) and present immense challenges due to inherent chemoresistance and poor outcomes. As TP53 ^mut are more common in older individuals and those with secondary/therapy‐related myeloid neoplasms (MN), their incidence is expected to increase with an aging population and rising proportion of cancer survivors. Treatments used for other MN—intensive chemotherapy, hypomethylating agents, and the BCL‐2 inhibitor venetoclax—do not improve the survival of TP53 ^mut MN patients meaningfully. Additionally, further development of many promising agents has been discontinued, highlighting the challenges. Widespread acknowledgment of these problems led to the recognition of TP53 ^mut MN as a distinct entity in the 5th edition of the World Health Organization and International Consensus Classifications. However, critical discrepancies between the two classifications may lead to under‐ or overestimation of the prognostic risk. Here, we review recent advances in the biology, diagnosis, and treatment of TP53 ^mut MN. The development of TP53 ^mut MN is positioned at the intersection of age, hereditary predisposition, and anti‐cancer therapies. Precursor TP53 ^mut clones can be detected years prior to the eventual leukemic transformation—raising the possibility of early intervention. We discuss the two classification systems and the bearing of the discrepancies between the two on timely and effective management. We provide novel evidence in the areas of discrepancies. Finally, we review the current therapeutic landscape and the obvious limitations of the currently used therapies.

1. Introduction

Alterations in tumor suppressor gene TP53—the “guardian of the genome”—are the most prevalent abnormalities in human cancers and are generally associated with poor outcomes [1]. TP53 alterations are relatively uncommon in myeloid neoplasms (MN), accounting for 8%–10% of myelodysplastic syndrome (MDS) and 10%–12% of acute myeloid leukemia (AML) cases [2, 3, 4, 5, 6]. Despite that, it remains a vexing challenge for three principal reasons: first, it is a highly aggressive leukemia with median survival of less than 1 year. Second, rapid advances in other types of MN have failed to translate into improved survival for this subset [7]. Effective treatments of TP53 ^mut MN remain elusive, primarily due to the inherent chemorefractory nature [8, 9, 10]. Commonly employed treatments including intensive chemotherapies or hypomethylating agents (HMA) confer little benefit. While selective targeting of BCL‐2 with the BH3‐mimetic venetoclax has been revolutionary for other types of AML, it does not translate in meaningful improvement in survival in this subset [11, 12]. Moreover, the development of multiple novel agents aimed at TP53 ^mut MN have been discontinued due to lack of efficacy [13]. Finally, TP53 ^mut are highly enriched in older patients as well as those with secondary and therapy‐related MN (t‐MN) [14]. Therefore, with aging population and the success of anti‐cancer therapies, the incidence of TP53 ^mut MN is expected to rise. Collectively, improving outcomes of TP53 ^mut MN remains one of the greatest unmet challenges.

Wide acknowledgement of these challenges led to the recognition of MN harboring TP53 ^mut as a separate entity by the 5th edition of the World Health Organization (WHO‐5) as well as International Consensus Classifications (ICC). The stated goals for the distinct grouping include the wider recognition of extremely poor prognosis, encouraging research, and facilitating clinical trial design—ultimately stimulating drug discovery. However, wide adoption has been challenging given critical differences between the guidelines.

The objectives of this review are to summarize advances in biology, diagnostic criteria, classification, and treatment options for TP53 ^mut MN. We discuss the known precursor lesions of TP53 ^mut MN—namely clonal hematopoiesis (CH) and clonal cytopenia of undetermined significance (CCUS)—and factors associated with leukemic transformation, though admittedly, the knowledge is rapidly evolving. Our special emphasis is on the challenges brought on by discrepancies in the diagnostic classifications and on reviewing our approach to diagnosis and management. Finally, we review the current therapeutic landscape and its obvious limitations.

1.1. Biology of TP53

The tumor protein p53, encoded by the TP53 gene located on chromosome 17p13, is a 53‐kilodalton protein that was initially identified in virally transformed cells and was categorized as an oncogene. However, subsequent research established that wild‐type p53 functions as a tumor suppressor, inhibiting growth and oncogenic transformation [15].

TP53 belongs to an evolutionarily highly conserved family of transcription factors: TP53, TP63, and TP73. The wild‐type p53 protein consists of 393 amino acids and contains several functional domains: two N‐terminal transactivation domains, a conserved proline‐rich domain, a central DNA‐binding domain (DBD), and a C‐terminus encoding its nuclear localization signals and an oligomerization domain required for the transcriptional activity.

1.1.1. p53: A Master Regulator of Diverse Cellular Processes

p53 has been implicated in a wide array of biological processes, including cell cycle arrest, senescence, apoptosis, autophagy, metabolism, and aging. It is also critical for maintaining genomic stability by balancing cell growth and cell arrest during genomic stress. p53 is usually present at a very low level in cells, and its short half‐life is regulated by post‐translational ubiquitination, acetylation, and phosphorylation. In a non‐stressed condition, p53 is ubiquitinated by the ubiquitin E3 ligase mouse double minute‐2 homolog (MDM2) leading to its proteasome‐mediated degradation [1]. A wide range of cellular stressors, including oncogenic, hypoxic, or DNA damage, inhibits MDM2, inhibiting ubiquitination of p53 and stabilizing it in tetrameric form. The tetrameric p53 induces the transcription of a diverse set of genes, turning on a highly complex anti‐proliferative transcriptional program that regulates virtually all known cellular functions (Figure 1A) [1, 16, 17, 18, 19]. This central location of the “TP53‐MDM2 axis”—in part—explains why TP53 alterations are the most common abnormalities in cancers [1]. p53 also plays a crucial role in maintaining genomic integrity, and regulating quiescence, self‐renewal, and differentiation of hematopoietic stem cells (HSC)—protecting against leukemogenesis [20, 21].

FIGURE 1.

Biology of TP53 loss in cancers. (A) Consequences of the loss of TP53 in carcinogenesis. In the inner circle, red denotes suppressive effect, orange denotes mixed effect, and green denotes promoting effects; (B) Mechanisms of TP53 inactivation; and (C) Common mechanisms of mono‐ and biallelic TP53 inactivation in myeloid neoplasms (see text for the studies used to determine prevalence).

TP53 ^mut affects p53's tetrameric conformations, impairing its ability to bind to transcriptional targets [22]. Hence, missense mutations along with truncating mutations and/or loss‐of‐heterozygosity/copy neutral loss‐of‐heterozygosity (LOH/cnLOH) can lead to loss of function, causing an inability to trigger p21, downregulation of genes associated with apoptosis, and upregulation of proteins involved in cell‐cycle progression and those involved in DNA‐damage repair (DDR). Murine studies also reported gain‐of‐function of missense mutations by demonstrating neomorphic protein–protein interactions of mutant p53 with other transcription factors. Moreover, a relative competitive fitness advantage of HSCs carrying missense mutations in the DBD over HSCs with monoallelic TP53 inactivation was observed [23]. Finally, missense TP53 ^mut can also exert a dominant‐negative effect on the wild‐type protein by forming tetramers with wild‐type p53, thus disrupting its transcriptional activity [23].

1.1.2. Mechanisms of TP53 Inactivation

While TP53 ^mut is the most common mechanism of inactivation, other mechanisms include the TP53 deletion, alternative splicing, post‐translational modification, micro‐RNA mediated degradation, or MDM2/MDM4 upregulation, ARF downregulation (Figure 1B). Two‐thirds of TP53 ^mut lead to biallelic inactivation either due to biallelic mutations or a single mutation with LOH/cnLOH of the wild‐type allele (Figure 1C). In MN, the mechanisms of biallelic TP53 inactivation include: ≥ 2 mutations in 18.4%–28.8% (median 22.6%), 1 mutation with deletion of the trans allele in 22.5%–42.2% (median 33%), and 1 mutation with concomitant cnLOH in 12.2%–20.6% (median 17%) of patients [8, 24, 25, 26]. (and Shah et al., under review).

2. Pathogenesis of TP53‐Mutated Myeloid Neoplasms

Relative infrequency of t‐MN and long latency from cytotoxic therapies suggest the role of factors beyond CH in leukemic transformation. The incidence of t‐MN is 0.5%–8% following cytotoxic exposures for various cancers. Second, median latency from the diagnosis of primary cancer to t‐MN diagnosis is approximately 7 years, with 54% and 29% diagnosed ≥ 5 and ≥ 10 years from primary malignancy, respectively [27].

Recent advances in sequencing technology have revolutionized our understanding of the occurrence of clones in myeloid driver genes with preserved blood counts (CH) or clonal CCUS. The explosion in our understanding of these precursor lesions has provided critical insight into leukemogenesis, including the origin and leukemic transformation of TP53 ^mut clones.

That TP53 ^mut are early leukemogenic mutations is now well established. TP53 ^mut constitutes 4%–5% of all CH [28, 29, 30, 31] and paired deep‐sequencing at the time of TP53 ^mut t‐MN and at/before primary malignancy demonstrated identical TP53 ^mut clone in t‐MN in a subset of cases. A seminal study demonstrated that mutational burden in the genomic region containing TP53 was comparable between t‐AML and de novo AML, suggesting that it was unlikely that chemotherapy directly induced the TP53 ^mut [32]. Indeed, in a smaller subset, the eventual TP53 ^mut clone was detected even prior to the institution of any cytotoxic therapies. Combined, these observations strongly suggest preexistence of the clone that gained a fitness advantage under the selective pressure [32].

Later, larger studies confirmed these observations in the context of diverse primary malignancies and cytotoxic exposures [33, 34, 35, 36]. For example, in a case–control study of cancer patients treated with cytotoxic therapy, 36% of cases of t‐MN developed harbored TP53 ^mut. Paired sequencing of PB/BM samples at the time of primary cancer diagnosis showed somatic TP53 ^mut at VAF 0.92%–22.3% in all evaluable cases [36]. Similarly, lymphoma patients undergoing autoHCT who later developed t‐MN, all 4 TP53 ^mut MN cases had pre‐existing TP53 ^mut clones at a low (0.9%–12%) VAF prior to autoHCT [33].

A study of 5978 patients with nonhematological cancers showed a 2.8‐fold higher risk of TP53 ^mut CH in patients who received cytotoxic therapies. The risk differed based on the type of therapies: platinum (2.1‐fold), radiation therapy (1.8‐fold) and taxane (1.9‐fold). Another study confirmed some of these findings: chemotherapy—particularly platinum and topoisomerase II inhibitors—preferentially selected for mutations in DDR genes, including TP53 [33, 35]. Radiation therapy was associated with the selective growth of TP53 ^mut and in those developing TP53 ^mut t‐MN, the TP53 ^mut clone was the dominant clone, suggesting clonal selection of TP53 ^mut under the strong cytotoxic pressure [35].

Thus, while it is clear that at least a subset of TP53 ^mut MN is an end‐result clonal expansion/evolution of the preexistent TP53 ^mut clone. However, the presence of TP53 ^mut is necessary and sufficient for the development of TP53 ^mut MN is not well understood. For example, in contrast to the cancer studies presented above, a population‐based study of 438 890 participants showed that TP53 ^mut CH/CCUS did not increase the risk of subsequent MN (HR 0.94, p = 0.87). In murine studies, TP53 ^mut HSC can promote self‐renewal but failed to induce overt transformation into leukemia, indicating that the mere presence of the TP53 ^mut clone may be insufficient to initiate leukemia [20, 37] and additional selection pressure may be necessary for clonal expansion/evolution and transformation. Therefore, HSC‐intrinsic factors discussed above, including the cooperation of BRCA1/2 PGV with TP53 ^mut and the haploinsufficiency of genes mapped on the minimally deleted region on chromosome 5q [35]. In addition, a single‐cell multi‐omics study aptly demonstrated that a chronic inflammatory microenvironment confers a fitness advantage and promotes evolution of TP53 ^mut while suppressing TP53 ^wt HSC [38].

Collectively, emerging evidence suggests a diverse risk of leukemic transformation of TP53 ^mut clones and that cooperation of HSC‐intrinsic factors such as genetic predisposition with HSC‐extrinsic factors such as genotoxic therapies and the bone marrow microenvironment can potentially shape clonal evolution and expansion. Factors associated with leukemic transformation of precursor states—CH and CCUS—are an area of active research.

2.1. TP53 Mutations Are Associated With Immunosuppressive Milieu

In addition to the immense impacts on the leukemic cells, TP53 ^mut modulates diverse aspects of the innate and adoptive immune system—further contributing to leukemogenesis. An elegant study showed that TP53 ^mut clones reshape the microenvironment conducive to their survival, chemoresistance, and immune evasion. TP53 ^mut MN HSCs have a higher expression of programmed death ligand‐1 (PD‐L1), which in turn is associated with c‐Myc upregulation and miR‐34a downregulation. Furthermore, the bone marrow of TP53 ^mut MN harbors a reduced number of cytotoxic and helper T‐cells and natural killer (NK) cells, but shows expansion of highly immunosuppressive regulatory T‐cells (T_regs) and myeloid‐derived suppressor cells—all suggestive of a highly immunosuppressed BM milieu [39]. TP53 ^mut AML were associated with increased numbers of activated B‐cells, effector memory CD4⁺ T‐cells, central memory CD8⁺ T‐cells, and two NK‐cell‐rich clusters [40].

Collectively, these results suggest a profound immune dysregulation, with features of immune senescence and an overall immune evasive phenotype which could be potentially leveraged to develop immunotherapy for TP53 ^mut MN. However, modest responses to immune checkpoint inhibitors (ICI) [41, 42, 43], and recent failures of anti‐CD47 monoclonal antibody [44], as well as monoclonal antibody against T‐cell Immunoglobulin Mucin (TIM)‐3 (sabatolimab) [45] suggest that overcoming the immunosuppressive milieu is a critical impediment to developing successful therapies (Table 1).

TABLE 1.

Clinical and laboratory studies of TP53‐mutated (TP53 ^mut) myeloid neoplasms (MN).

Characteristics	Study
	Bahaj [25]	Grob a [8]	Haase b [10]	Kaur c [46]	Shah d	Weinberg e [47]
N	1010	230	186	173	580	247
Age in years, median	71	62	70	67.9	68.6	70
Male (n, %)	536 (53%)	136 (59%)	108 (58%)	104 (60%)	366 (63%)	129 (52%)
BM Blasts < 5% (n, %)	NR	0 (0%)	54 (29%)	0 (0%)	194 (33.4%)	43 (17.4%)
BM blasts 5%–9% (n, %)	NR	44 (19.1%)	59 (31.7%)	0 (0%)	75 (12.9%)	37 (15.0%)
BM blasts 10%–19% (n, %)	NR		65 (34.9%)	36 (20.8%)	92 (15.9%)	54 (21.9%)
BM blasts ≥ 20% (n, %)	NR	186 (80.9%)	3 (1.6%)	137 (79.2%)	219 (31.8%)	113 (45.7%)
Prior therapy (%)	NR	NR	NR	NR	278 (47.9%)	106 (42.9%)
Hemoglobin (g/dL), median	9.1	NR	9.2	7.9	8.9	8.5
WBC (109/L), median	5.1	NR	NR	NR	3	3
Platelets (109/L), median	65	NR	47	40	55	50
Peripheral blasts %	NR	NR	NR	—	1	1

Abbreviations: ANC—absolute neutrophil count; BM—bone marrow; WBC—white blood cells.

^a MDS‐EB/AML cases only.

^b MDS with CK only.

^c Multihit TP53 ^mut cases with blasts ≥ 10% only.

^d In revision.

^e MN with CK only.

3. Epidemiology of TP53‐Mutated Myeloid Neoplasms

The Pan‐Cancer cohort confirmed TP53 as the most commonly altered gene across various cancers. The prevalence of its alterations varies widely—from ~95% in serous ovarian cancers to ~2% in renal cancers [4]. MN are positioned at the lower end of this spectrum, with 7.5% of MDS and AML harboring TP53 ^mut [3, 4]. However, TP53 ^mut is enriched in 3 scenarios in MN: (1) older individuals; (2) individuals with germline pathogenic variants (PGV) including Li‐Fraumeni syndrome, BRCA1, and BRCA2 carriers; and (3) patients who had prior cytotoxic or immunosuppressive therapies.

1. Older age. The median ages of TP53 ^mut AML and MDS are 67 and 73 years, respectively. TP53 ^mut AML is significantly more common in patients ≥ 60 compared to < 60 years of age (7% vs. 2%) and within the ≥ 60 cohort, the frequency of TP53 ^mut AML increases with age: 16%, 19%, 17%, and 50% in 61–70, 71–80, 81–90, and 91–100 years, respectively. On the other hand, no such difference was seen for MDS [14].

2. Pathogenic germline variant carriers. TP53‐deficient tumors may be enriched in patients with PGV—especially those in the DDR genes. It is thought that TP53 dysfunction is integral to BRCA1/2‐associated tumorigenesis, which could explain the enrichment of TP53 ^mut t‐MN in BRCA1/2 PGV carriers. The enrichment is likely heightened by the use of cytotoxic therapies for treating primary cancer in carriers of DDR genes [48]. TP53 ^mut t‐MN are indeed more prevalent in BRCA1/2 PGV carriers of breast and non‐breast/ovarian cancer survivors [49, 50]. Ovarian cancer patients with BRCA1/2 PGV are 6.2‐fold more likely to harbor TP53 ^mut CH compared to BRCA1/2 wild‐type individuals [51]. At least a subset of pre‐existing TP53 ^mut clones expanded during poly (ADP‐ribose) polymerase inhibitors (PARPi) therapy, suggesting a potential mechanistic link [35]. A study of 53 t‐MN patients that included 12 patients with prior breast or ovarian cancers found that approximately half harbored PGV in DDR genes [52].

Along the same lines, somatic inactivation of the trans allele—via the LOH, cnLOH, or copy gain LOH (cgLOH)—is a common occurrence in patients with Li–Fraumeni syndrome (LFS)—rare autosomal dominant cancer predisposition syndrome caused by PGV in TP53. LFS patients are at an increased risk of developing multiple cancers, including ~4% incidence of hematological malignancies [53]. In tumors arising in LFS individuals, LOH/cnLOH of TP53 was seen in a majority (86%) of cases that appeared to have occurred as early as in utero [54].

• 3Prior cytotoxic and immunosuppressive therapies. Cancer survivors are at a 4.7‐fold higher risk of developing leukemia compared to the general population. As the effectiveness of cancer‐directed therapies has improved survival, long‐term complications of these therapies have come to focus. For instance, the incidence of t‐MN has risen over the last two decades: from 0.04/10 000 new cases between 2001 and 2007 to 0.20/100 000 new cases between 2008 and 2014 [55].

TP53 ^mut is enriched in t‐MN (13%–48%) compared to de novo MN (2%–18%) (average 31% vs. 8%, Figure 2A, Table 1) [2, 3, 4, 6, 8, 9]. Solid and hematological cancers precede t‐MN in approximately 40% and 50%, respectively; the remaining develop following treatment for autoimmune rheumatological diseases (AIRD) [58] or solid organ transplant. t‐MN developing in MM and ovarian cancer patients is significantly enriched with TP53 ^mut [59]. Among treatments, autologous hematopoietic stem cell transplant (autoHCT) and chemotherapy exposure were associated with 2.2‐ and 2‐fold higher risk of TP53 ^mut t‐MN compared to radiation exposure alone [27].

FIGURE 2.

Molecular and chromosomal abnormalities in TP53‐mutated (TP53 ^mut) myeloid neoplasms (MN). (A) TP53 ^mut is enriched in therapy‐related compared to de novo MN; (B) Lollipop Plot of TP53 ^mut in therapy‐related myeloid MN (t‐MN, 253 variants). Missense mutations are predominant in TP53 ^mut MN, with 85% of cases harboring one or more missense mutations. As with other cancers, > 80% of TP53 ^mut in MN localize to the DNA‐binding domain (DBD, 100–300 amino acids) and localize to 6 hotspot sites (R175, H179, Y220, G245, R248, R273); (C) OncoPlot for TP53 ^mut t‐MN (N = 182 patients). TP53 ^mut MN are associated with a paucity of co‐mutations in other myeloid driver genes and a very high prevalence of chromosomal abnormalities; (D) Percentage cases with co‐mutations in the most commonly co‐mutated genes; (E) Percentage cases with chromosomal abnormalities; and (F) In myeloid neoplasms, TP53 ^mut is enriched in complex karyotype. AID—autoimmune disease; chemo—chemotherapy; chr. abn.—chromosomal abnormality; del.—deletion; Hem—hematological malignancy; IS—immunosuppression; RT—radiotherapy; solid—solid tumor; t‐AML—therapy‐related acute myeloid leukemia; t‐MDS—therapy‐related myelodysplastic syndrome; VAF—variance allele frequency; WT—wild‐type. Panel (A) source data from Singhal et al. [56] Panels (B), (C), and (F) are adapted from Shah et al. [57] In panels (D) and (E), each diamond represents a study, whereas the line (—) represents the weighted average across the studies when available.

In summary, emerging data suggest the propensity to develop TP53 ^mut clones due to aging or hereditary predisposition, a transformation that is accelerated by cooperating genetic defects or iatrogenic selection pressure.

3.1. Emerging Therapies Associated With an Increased Risk of TP53 ^mut MN

With increased utilization of targeted and immunotherapies, it was hoped that the incidence of t‐MN—and TP53 ^mut MN—would decrease. Emerging data, however, suggest that many novel therapies are associated with an increased risk of both. The following section examines the evolving connection between non‐traditional anticancer therapies such as peptide receptor radionuclide therapies (PRRT), PARPi, and chimeric antigen receptor (CAR) T‐cell therapies with TP53 ^mut MN and the commonalities.

Studying these connections is critical for two reasons: utilization of these therapies is expected to grow exponentially both in the breadth of indications as well as in volume. Second, t‐MN developing after the novel therapies shares some features including the proportion of TP53 ^mut MN, chromosomal abnormalities, and outcomes comparable to the traditional cytotoxic therapies, whereas latency appears to be substantially shorter—deserving additional studies.

Peptide receptor radionuclide therapies (PRRT) such as Lutetium‐177‐Dotatate (¹⁷⁷Lu‐Dotatate) and Lu 177 vipivotide tetraxetan (¹⁷⁷Lu‐PSMA‐617) are targeted systemic radiopharmaceutical therapies that use beta particle‐emitting radionuclides linked to peptide analogs. ¹⁷⁷Lu‐Dotatate is approved by the US Food and Drug Administration (FDA) for the treatment of somatostatin gastroenteropancreatic neuroendocrine tumors (GEP‐NETs). ¹⁷⁷Lu‐PSMA‐617 is approved for the treatment of metastatic castration‐resistant prostate cancer (mCRPC). Approximately 1.4%–4.8% of patients treated with ¹⁷⁷Lu‐Dotatate developed t‐MN, following a median latency of 3–4 years from the treatment [60, 61]. In a recent study describing 14 t‐MN following PRRT therapies, 2 multi‐hit TP53 ^mut MDS developed at 17 and 18 months following the therapy. In addition, two patients had mutations in protein phosphatase magnesium‐dependent (PPM)‐1D gene, which were mutually exclusive [62].

A recent single‐center case series described 5 (1.3%) t‐MN developments among 381 (1.3%) patients treated with ¹⁷⁷Lu‐PSMA, of which 3 had multi‐hit TP53 inactivation and high‐risk karyotype in all cases [63].

PARPi are the standard‐of‐care therapy for ovarian, breast, or metastatic prostate cancer with known or suspected defective DDR mechanisms. PARP is involved in the repair of single‐strand DNA breaks, while DDR genes such as BRCA1 and BRCA2 facilitate double‐strand DNA breaks. In a study of 298 BRCA1/2 mutated cancer patients treated with olaparib, 6 (2%) t‐MN were noted [64]. A meta‐analysis of 28 randomized controlled trials comparing PARPi‐ (n = 5693) and placebo‐treated (n = 3406) patients showed that PARPi was associated with a 2.6‐fold higher risk of t‐MN (0.73% vs. 0.47%). Notable features included a short duration of PARPi exposure (median 9.8 months) and a strikingly short latency to develop t‐MN (median 20.3 months) following PARPi exposure [65]. Between 50%–75% of t‐MN following PARPi had TP53 ^mut, the majority being multi‐hit or multi‐hit equivalent [66, 67]. Enrichment of TP53 ^mut t‐MN could partially be ascribed to a higher burden of TP53 ^mut CH (64% vs. 14.3%) in PARPi‐treated patients compared to those who did not receive PARPi as well as clonal expansion during PARPi therapy [67]. Taken with the knowledge that a proportion of these patients carry cancer susceptibility PGV (discussed above), this may explain the high incidence of TP53 ^mut t‐MN in this group.

Chimeric antigen receptor (CAR) T‐cell therapies have revolutionized the management of relapsed or refractory B‐cell acute lymphoblastic leukemia, CD19+ LPD, and MM. With wider utilization and longer follow‐up, an increased risk of secondary cancers—including t‐MN—is emerging and has been an area of immense focus recently [68, 69, 70, 71, 72, 73, 74, 75].

The cumulative incidence of t‐MN following CAR‐T therapy (n = 312) was 4%, 6%, and 9% at 1‐, 2‐, and 3‐years, respectively. Notably, 44.4% of t‐MN cases harbor TP53 ^mut. A striking feature of post‐CART t‐MN is the short latency (median 9.1 months from CART therapy) with 60% of t‐MN following CAR‐T [69] diagnosed within 1 year. This is in stark contrast to the much longer latency of 5–6 years following conventional cytotoxic therapies including autoHCT [33, 48, 76]. The short latency and high incidence of TP53 ^mut following CAR‐T therapy were comparable between LPD and MM cases treated with CD19‐ and BCMA‐directed constructs—suggesting a potential mechanistic link between CAR‐T therapy and the subsequent MN. A higher proportion of TP53 ^mut t‐MN can partly be explained by the high prevalence (37%–64%) of CH at baseline, predominantly involving TP53 and PPM1D [70, 72]. The same TP53 ^mut clone was detected in a subset of post CAR‐T t‐MN, suggesting that CAR‐T therapy may lead to rapid expansion of pre‐existing clones via hitherto unexplained mechanisms [70].

In summary, the study of t‐MN developing after novel therapies is an area of immense interest. Whereas comparable genetic and genomic features (including the proportion of multi‐hit TP53 ^mut) strongly suggest shared pathogenesis, these novel modalities are used in conjunction with or following cytotoxic therapies. Therefore, whether the above therapies induce mutations or cause the expansion of the pre‐existing mutation is unclear. Further research is needed to isolate the impact of these therapies on TP53 ^mut MN development and the mechanisms thereof.

4. Clinical and Molecular Characteristics of TP53‐Mutated Myeloid Neoplasms

4.1. The Landscape of TP53 Mutations in MN

Deletion of the entire chromosome 17 or 17p13.1 across the TP53 locus and truncating mutations can lead to TP53 loss of function. However, unlike for most other suppressor genes, missense mutations are predominant in TP53 ^mut MN, with 77% being missense, 9% frameshift, 6% splice, 6% nonsense, 2% deletion, and 1% silent variants (Figure 2B). As with other cancers, > 80% of TP53 ^mut in MN localize to the DNA‐binding domain (DBD, 100–300 amino acids) and localize to hotspot sites [77]. MN, including t‐MN, harbor a distinct profile of mutational hotspots when compared to other cancers. An analysis of the International Agency for Research on Cancer TP53 database showed that 6 commonest mutations (R175, G245, R248, R249, R273, and R282) account for 27.7% of all TP53 ^mut across cancers [77]. In contrast, the mutational spectrum of all MN and t‐MN showed a consistently distinct pattern, with R175, Y220, M237, R248, R273, and R282 being the most common variants [25, 57]. The biological basis and impact of the mutational characteristics (e.g., location, hot spot vs. not) in MN remain poorly understood.

4.2. The Pattern of Co‐Mutations in TP53‐Mutated MN

Given the well‐established role of TP53 in DDR, it may be assumed that TP53 ^mut MN would have a higher mutation burden. However, a striking feature of TP53 ^mut MN compared to TP53 ^wt MN is the paucity of co‐mutations in myeloid driver genes [57, 78]. More than half (52%) of TP53 ^mut t‐MN harbor no known co‐mutations compared to 17.3% of TP53 ^wt t‐MN [9] with an average number of co‐mutations being lower in biallelic TP53 ^mut compared to TP53 ^wt MN [25] (0.8 vs. 2.1 per case, Figure 2C,D). In TP53 ^mut t‐MN, ≥ 2 co‐mutations were noted in only 20.3% of TP53 ^mut cases compared to 63.7% of TP53 ^wt t‐MN [9]. The paucity of co‐mutations is across the genes/pathways including signaling pathway (KIT, FLT3, WT1), epigenetic modifiers (DNMT3A, TET2, or ASXL1), or spliceosome (U2AF1, SF3B1, SRSF2, or ZRSR2). On the other hand, alterations in ETV6 and NF1 are enriched in TP53 ^mut MN [78]. Even among the MN harboring TP53 ^mut, “TP53‐driven” tumors (i.e., multi‐hit TP53 ^mut) were characterized by a paucity of co‐mutations compared to single‐hit TP53 ^mut [24, 57], suggesting that multi‐hit TP53 inactivation is a biologically distinct entity compared to both single‐hit TP53 ^mut and TP53 ^wt MN.

In the absence of a clear co‐mutational pattern associated with survival in TP53 ^mut MN, scoring systems such as Evolutionary action for TP53 (EAp53) [79, 80, 81] and EPI‐6 scores [46] have been proposed. EAp53 is a computational approach initially developed for head and neck squamous cancers that has been applied to TP53 ^mut MN with variable success [80, 81]. Kaur et al. recently proposed the EPI‐6 score that accounts for alterations in one or more of the 6 genes (CUX1, U2AF1, EZH2, TET2, CBL, or KRAS) that predicted inferior 2‐year survival in venetoclax and HMA‐treated patients [46]. The lack of independent validation has limited wider adoption of these scoring systems in routine practice.

4.3. Chromosomal Abnormalities in TP53‐Mutated MN

MN are remarkably “silent” genomically compared to other cancers [4, 82], and chromosomal abnormalities are noted in only 40%–50% of MN [3, 25, 83, 84]. Given the role of TP53 as “the guardian of the genome,” it is conceivable its loss is associated with genomic alterations. In the absence of functional p53, cells are unable to effectively repair DNA‐damage—culminating into large chromosomal alterations. Indeed, a striking feature of TP53 ^mut MN is the pervasiveness of cytogenetic abnormalities including highly complex genomic aberrations and chromothripsis [24, 57].

Metaphase cytogenetic studies demonstrate the enrichment of complex karyotypic (CK; defined as ≥ 3 chromosomal abnormalities in the absence of AML‐defining translocations and/or inversions) abnormalities in 70%–80% of TP53 ^mut MN compared to 10% of TP53 ^wt MN (Figure 2E) [25]. Similarly, chromosomal abnormalities are more prevalent in TP53 ^mut t‐MN compared to TP53 ^wt t‐MN—50%–60% of TP53 ^wt t‐MN harbor a normal karyotype, whereas only 12.4% of TP53 ^mut t‐MN have a normal karyotype. Recurrent cytogenetic abnormalities, including deletion 5q, deletion 7q, and loss of 17p, are all common in TP53 ^mut MN. Conversely, the frequency of TP53 ^mut increases from 4.5% in normal karyotype cases to 17.3% in the presence of two chromosomal aberrancies and 76.8% in the presence of CK. Even within the CK group, TP53 ^mut is further enriched with the increasing number of cytogenetic abnormalities: from 26.3% in cases with 3 chromosomal abnormalities to 75%, 96.6%, and 94% in cases with 4–6, 7–9, and > 9 chromosomal aberrancies (Figure 2F).

Finally, the association of CK on survival in TP53 ^mut is context dependent: in a cohort of de novo TP53 ^mut MN, the presence of CK was associated with inferior survival compared to non‐CK TP53 ^mut MN [8]. Conversely, in a cohort of 287 MN harboring CK, 83% harbored TP53 ^mut and were an independent risk factor for inferior survival [47]. The impact of CK appears to be limited to MDS with no increased blasts but not in cases with increased blasts [26]. These observations do not translate to TP53 ^mut t‐MN, wherein survival is poor regardless of CK (7.3 vs. 8.3 months) [57].

Finally, p53 suppresses chromosome shattering and rearrangement events known as chromothripsis. Chromothripsis is characterized by massive genomic rearrangements that are often generated in a single catastrophic event and an oscillating pattern of DNA copy‐number levels in one or a few chromosomes [82, 85]. Consequently, the loss of p53 facilitates accumulation and permits the survival of aneuploid cells. Pan‐Cancer Analysis of Whole Genomes (PCAWG) across all tumor types showed that TP53 ^mut was associated with a 1.5‐fold higher risk of chromothripsis compared to TP53 ^wt (38% vs. 24%) [82]. In a study of AML with CK, chromothripsis was noted in 35% of cases, and chromosomes 7 (10%), 3 (9%), and 12 (9%) were the most commonly involved targets. Chromothripsis is independently associated with a 2.5‐fold higher risk of death in AML [85].

4.4. Impact of TP53 Mutations in MDS With Isolated Deletion of the Long Arm of Chromosome 5

MDS with isolated deletion of the long arm of chromosome 5 [MDS‐del(5q)] is a distinct sub‐entity of MDS characterized by superior outcomes, a favorable response to lenalidomide, and enrichment of SF3B1 mutations [86]. TP53 ^mut are enriched in MDS‐del(5q) cases (18%–20% vs. 5%–10% in other MDS) and are further enriched in the leukemic phase [87, 88, 89]. TP53 ^mut are also enriched in therapy‐related compared to de novo MDS‐del(5q) (30% vs. 19%) [89], but a significantly lower proportion harbors multi‐hit inactivation compared to non‐del(5q) MDS (25% vs. 70%–75%) [16, 88].

TP53 ^mut is associated with inferior survival and increased risk of AML progression in this otherwise favorable‐risk category and thus represents cases at a crossroad of the opposite ends of the prognostic spectrum [90]. It is thought that genes in the minimally deleted region on chromosome 5q, including casein kinase‐1 alpha 1 (CSNK1A1), Early Growth Response 1 (EGR1), and APC, cooperate with TP53 ^mut to confer a survival advantage in HSC and progression to AML [91, 92]. Inhibition of CSNK1A1 induces p53‐mediated apoptosis; selective lenalidomide‐mediated degradation of CSNK1A1 is more toxic to normal HSC, providing the competitive advantage seen for the TP53 ^mut clone [59].

Recent studies identified multi‐hit TP53 ^mut or monoallelic TP53 ^mut with VAF $\ge$ 20 [88] or $\ge$ 22% [93] as the subset with poor survival. In contrast, monoallelic TP53 ^mut cases with VAF < 22% had survival and AML progression risks comparable to TP53 ^wt MDS‐del(5q) [88]. While highly effective, lenalidomide therapy can lead to selective expansion of TP53 ^mut clones [59] raising the possibility of leukemic transformation.

In conclusion, integrating TP53 ^mut allelic status and VAF threshold into diagnostic, monitoring, and management strategies can better stratify risk and guide treatment decisions in MDS‐del(5q).

4.5. The Impact of Multihit TP53‐Mutations in Myeloproliferative Neoplasms

In the current form of WHO‐5 and ICC, TP53 ^mut MN does not include myeloproliferative neoplasms [86, 94]. A recent study evaluated the impact of TP53 ^mut (VAF ≥ 2%) in myeloproliferative neoplasm (MPN) cases. Accelerated/blast (AP/BP) cases harboring TP53 ^mut had a significantly shorter survival compared to TP53 ^wt MPN‐BP/AP, regardless of the allelic status or VAF. On the other hand—reminiscent of the TP53 ^mut MDS—multihit TP53 ^mut had a shorter survival compared to non‐multihit TP53 ^mut for chronic phase myelofibrosis [95].

5. Diagnosis of TP53‐Mutated Myeloid Neoplasms

TP53 ^mut is associated with poor survival in all MN including MDS, AML, MDS/MPN, and MPN [6, 8, 9, 96, 97] (and Tefferi et al., in print). Recently proposed the 5th edition of the WHO classification (WHO‐5) and International Consensus Classification (ICC) acknowledged TP53 ^mut MDS and AML as a distinct entity in recognition of the biological homogeneity, poor survival, and biological and clinical distinctness from TP53 ^wt MN [86, 98]. Regardless of the blast percentage—TP53 ^mut MN shares striking similarity in the enrichment of cytogenetic abnormalities, paucity of co‐conspiring mutations, and most importantly, outcomes [8, 9].

Classification of TP53 ^mut MN can be seen as a two‐step process: First, the assignment of TP53 hit status and second, the integration of the hit status with blast % category. TP53 ^mut is usually established by targeted sequencing analysis covering at least exons 4–11. Molecular pathology methods capable of detecting LOH include fluorescence in situ hybridization (FISH), single nucleotide polymorphism (SNP) array, a specialized NGS panel designed to detect 17p LOH, and WGS. cnLOH can be detected by SNP array, the specialized NGS panel, or WGS. The presence of ≥ 2 mutations usually targets both alleles and is considered biallelic/multi‐hit events. Similarly, the presence of one TP53 ^mut with VAF ≥ 50% is considered to be presumptive (not definitive) evidence of LOH/cnLOH (Table 2).

TABLE 2.

Summary of the diagnostic criteria for myeloid neoplasms harboring TP53 mutations (TP53 ^mut) between the 5th Edition of the World Health Organization (WHO‐5) and International Consensus Classifications (ICC).

Blast % category	WHO‐5	ICC	Comments
0%–9%	MDS with biallelic TP53 inactivation: ≥ 2 TP53 mut or 1 TP53 mut with TP53 copy number loss or cnLOH*	Multi‐hit: ≥ 2 TP53 mut (each with VAF ≥ 10%) or 1 TP53 mut (with VAF ≥ 10%) and (1) deletion involving the TP53 locus at 17p**; (2) VAF ≥ 50%; or (3) copy neutral LOH at the 17p. Multi‐hit equivalent: One TP53 mutation (VAF ≥ 10%) and a complex karyotype	Emphasis on “multi‐hit status” in both ICC and WHO‐5. ICC requires VAF ≥ 10%, WHO‐5 does not. In the context of 1 TP53 mut (VAF ≥ 10%) and the absence of LOH information, CK is considered “multihit equivalent” by ICC but not by WHO‐5
10%–19%		MDS/AML with mutated TP53: TP53 mut VAF ≥ 10%, regardless of allelic status	Continued emphasis on biallelic loss in WHO‐5, but not ICC. Continued emphasis on VAF ≥ 10% in ICC, but not WHO‐5.
≥ 20%	No distinct designation and included with other AML	AML with mutated TP53: TP53 mut VAF ≥ 10%, regardless of allelic status	No separate designation in WHO‐5. Continued emphasis on VAF ≥ 10% in ICC.

Abbreviations: AML—acute myeloid leukemia; BM—bone marrow; cnLOH—copy neutral LOH, LOH—loss‐of‐heterozygosity; MDS—myelodysplastic syndrome; VAF—variant allele frequency.

^*Per WHO‐5, copy number variation (CNV) analysis is required to confirm the loss of the TP53 locus, as the mere detection of the 17p13.1 deletion is not sufficient.

^**Per ICC, a cytogenetic deletion involving the TP53 locus at 17p13.1 is adequate and does not mandate confirmatory CNV analysis.

5.1. Review of the 5^th Edition of the WHO and International Consensus Classification for TP53‐Mutated Myeloid Neoplasms

The stated goals for distinct classification of TP53 ^mut MN include the wider recognition of extremely poor prognosis, encouraging research, and facilitating clinical trial design—ultimately stimulating drug discovery. In practice, however, the adoption of the classifications has been challenging due to discrepant criteria. For example, in a study of 603 MN harboring TP53 ^mut (VAF ≥ 2%), 64% and 20.4% of TP53 ^mut MN cases would not be classified as TP53 ^mut MDS or MN by WHO‐5 and ICC, respectively (Figure 3A). Moreover, 407 (67.5%) would be classified discrepantly (Figure 3B) [99]. Another study (n = 188) confirmed these findings to demonstrate that 64% of cases would be classified differently by WHO‐5 and ICC [100]. As both classifications are increasingly used and governing clinical practice, the differences between the two classifications can lead to either under‐ or overestimation of the prognostic risk and inconsistencies in treatment decisions.

FIGURE 3.

Myeloid neoplasms (MN) harboring TP53 mutations (TP53 ^mut) are classified discrepantly using the 5th edition of the World Health Organization (WHO‐5) and International Consensus Classifications (ICC). (A) Schema representing the 5th edition of the WHO (WHO‐5, for MDS with biTP53) and International Consensus Classifications (ICC, myeloid neoplasms with mutated TP53); (B) Retrospective application of MN harboring TP53 ^mut (variant allele frequency ≥ 2%) classifies a 67.5% of the cohort discrepantly; (C) interaction between blasts, allelic status, complex karyotype, and VAF. AEL—acute erythroleukemia; AML—acute myeloid leukemia; AML‐MR—AML myelodysplasia related; CCUS—clonal cytopenia of undetermined significance; CK—complex karyotype; EB—excess blasts; EB—excess blasts; LB—low (< 5% blasts); MDS—myelodysplastic syndrome; MLD—multilineage dysplasia; MN—myeloid neoplasms; PEL—pure erythroid leukemia; SLD—single‐lineage dysplasia; VAF—variant allele frequency; VAF—variant allele frequency. Panel B is adapted from Shah et al. (under review).

The critical differences between the two classifications are driven by the assigned importance to morphological assessment, TP53 ^mut AML, allelic status, and blast percentage in cases with MDS, complex karyotype, 17p13.1 deletion detected on metaphase cytogenetic, and TP53 ^mut VAF threshold (Table 3). The WHO‐5 classification includes only biallelic TP53 ^mut MDS, whereas the ICC includes MDS and AML in a distinct category of TP53 ^mut MN.

TABLE 3.

Areas of discrepancy between the 5th edition of the World Health Organization (WHO‐5) and International Consensus Classifications (ICC) for the classification of TP53‐mutated (TP53 ^mut) myeloid neoplasms (MN).

No.	Area of discrepancy	WHO‐5: MDS with biallelic TP53 inactivation	ICC: Myeloid neoplasms with mutated TP53	Emerging evidence	Conclusions
1	Acute myeloid leukemia (AML)	No separate category, most cases would be classified as AML‐MR.	Separate category (BM/PB blasts ≥ 20%), includes PEL	Survival of TP53 mut AML is significantly poor compared to TP53 wt AML‐MR (4.7 vs. 18.3 mo.; p < 0.0001) [99, 100].	TP53 mut AML should be included in TP53 mut MN.
2	Variant allele frequency (VAF) threshold	None	VAF ≥ 10%	In cases with TP53 mut VAF < 10%: Survival of AML and multi‐hit TP53 mut “MDS” was significantly poor compared to their counterpart with single hit, and comparable to cases with VAF ≥ 10% (14.1 vs. 48.8 vs. 7.8 mo., p < 0.0001).	“Multihit” TP53 mut with VAF 2% to < 10% should be included in TP53 mut MN.
3	Interaction of blast % categories with allelic status	Requires demonstration of biallelic inactivation for blasts 0%–19%	Requires demonstration of multi‐hit status for blasts 0%–9%	Survival of monoallelic MDS ≥ 5% is comparable to biallelic inactivation [99] and (Shah et al., under review).	Monoallelic TP53 mut MDS with blasts ≥ 5% should be included in TP53 mut MN.
4	Complex karyotype (CK) as multihit equivalent	CK is not considered a multi‐hit equivalent	CK is considered multi‐hit equivalent	Survival of single TP53 mut VAF < 50% with CK is comparable to 17p deletion on karyotype (10.4 vs. 11.0 mo.; p = 0.39) and poorer than monoallelic TP53 mut without 17p loss or CK (33.4 mo.; p < 0.0001) [99].	In single TP53 mut with VAF < 50%, CK should be considered multi‐hit equivalent.
5	Confirmation of 17p13.1 deletion detected on karyotype by CNV analysis	Requires confirmation of loss of TP53 locus detected on karyotype with an additional CNV method	Does not require confirmation of 17p13.1 deletion	Cases with single TP53 mut (VAF < 50%) and 17p13.1 deletion on karyotype: CNV analysis verified LOH across the TP53 locus in 94% cases [99]. Cases with single TP53 mut (VAF < 50%) without 17p13.1 deletion on karyotype: CNV analysis identified LOH/cnLOH in 26.9% of evaluable cases, all of which harbored CK [99].	In single TP53 mut with VAF < 50%, confirmation of 17p deletion with an additional CNV may not be necessary except complex rearrangements and structural changes
6	Morphology	Requires MDS diagnosis	Blasts < 5%: Cytopenia, does not require diagnostic dysplasia. Blast 5%–19%: Any cytopenia ≥ 20%/PEL: no requirement for cytopenia	—	Additional evidence needed
7	Prior cytotoxic therapies	Myeloid neoplasms post‐cytotoxic therapy (MN‐pCT), a part of “secondary myeloid neoplasms”	Diagnostic qualifier	In t‐MDS, survival of single‐hit TP53 is comparable to multi‐hit loss (10.2 vs. 9.7 mo.; p = 0.59) [9].	Single‐hit t‐MDS should be included in TP53 mut MN.

Abbreviations: AML—acute myeloid leukemia; AML‐MR—AML with myelodysplasia‐related; BM—bone marrow; cnLOH—copy neutral LOH; CNV—copy number variation analysis; LOH—loss‐of‐heterozygosity; MDS—myelodysplastic syndrome; MN‐pCT—myeloid neoplasms post‐cytotoxic therapy; mo.—months; PB—peripheral blood; PEL—pure erythroid leukemia; t‐MDS—therapy‐related myelodysplastic syndrome.

At the core of the above noted discrepancies is the ability to denote a neoplastic clone that completely lacks wild‐type TP53 function. Ideally, this would be established by demonstrating inactivation of both the alleles via ≥ 2 mutations or at ≥ 1 mutation with the deletion of the trans allele in the same cell. In the absence of routinely available single‐cell techniques to reliably establish biallelic inactivation, multi‐hit TP53 inactivation is considered to be an acceptable surrogate. Urgent efforts are underway to provide additional clarifications in the matter and will help inform future iterations of the guidelines [9, 25, 26]. Below we review the emerging evidence that is expected to provide clarifications in the areas of discrepancy.

1. TP53‐mutated AML. Approximately 35%–40% of TP53 ^mut MN are AML. A vast majority (94%) of these are discrepantly diagnosed between the ICC and WHO‐5 since ICC recognizes AML with mutated TP53 ^mut VAF $\ge$ 10%, whereas a vast majority would instead be classified as AML, myelodysplasia‐related (AML‐MR) by WHO‐5 [101]. However, TP53 ^mut AML showed a distinct genetic profile and significantly worse overall survival [99, 100]. Survival of TP53 ^mut AML was significantly poor compared to TP53 ^wt AML‐MR with myelodysplasia related (4.7 vs. 18.3 months; p < 0.0001) [99]. Collectively, these results support distinguishing TP53 ^mut AML from other AML and incorporating in the TP53 ^mut MN category [99, 100].

2. Variant allele frequency threshold. Another critical difference between the two classifications is the adoption of a VAF threshold of ≥ 10% in the ICC, but not WHO‐5. Specifically, ICC mandates TP53 ^mut VAF ≥ 10% with an implicit rationale of distinguishing it from smaller TP53 ^mut clones that may not be the pathologic driver. It is understood that the adopted VAF threshold is an empiric threshold to distinguish MN driven by TP53 ^mut from those harboring TP53 ^mut. Approximately 10% of TP53 ^mut MDS‐EB1, MDS‐EB2, and AML had TP53 ^mut VAF < 10% and 55.6% of these cases harbored CK. Survival of MDS‐EB1, EB2 with VAF < 10% without CK was comparable to monoallelic MDS‐LB without CK (26.2 vs. 34.8 months; p = 0.44), whereas survival of MDS‐EB1, EB2 with VAF < 10% with CK was comparable to their counterparts with VAF ≥ 10% (5.6 vs. 6.3 months). Collectively, these results demonstrate that TP53 ^mut MN cases with ≥ 5% blasts with CK should be included in TP53 ^mut MN regardless of the VAF ≥ 10% or < 10% [99].

3. Interaction of the blast percentage categories with allelic status. The two classifications differ in their interpretation of the interaction between the blast percentage categories and allelic/hit status. The ICC prioritizes blast percentage in risk stratification, proposing 3 subcategories: multi‐hit MDS (0%–9% BM/PB blasts), MDS/AML (10%–19% BM/PB blasts), and AML (≥ 20% BM/PB blasts) regardless of allelic status. Consequently, single‐hit MDS (0%–9% BM/PB blasts) is excluded from the definition of TP53 ^mut MN. On the other hand, WHO‐5 requires the demonstration of biallelic TP53 inactivation throughout the spectrum of MDS (blasts 0%–19%), thus excluding monoallelic TP53 ^mut MDS with 0%–19% blasts.

We have recently shown that the allelic status is of critical prognostic importance for cases with MDS‐low blasts (BM blast < 5% and blood blasts < 2%); but the prognosis of MDS‐EB1 (BM blasts 5%–9% and/or PB blast 2%–4%), MDS‐EB2 (BM blasts 10%–19% and/or PB blast 5%–19%), and AML (BM/PB blast ≥ 20%) is poor regardless of the allelic status. This is in agreement with another study demonstrating that MDS with blasts < 5% is a distinct subgroup compared to other categories. For example, only 24% of TP53 ^mut MDS with < 5% harbored multi‐hit TP53 inactivation compared to 67%, 91%, and 71% of cases with blasts 5%–9%, 10%–19%, and ≥ 20%, respectively [26]. Second, CK status predicted poor survival in MDS < 5% (hazard ratio, 5.2; p < 0.001) but not in the higher blast categories [26]. Combined, adopting WHO‐5 and ICC would underestimate the survival of single‐hit TP53 ^mut MDS 5%–19% and 10%–19%, respectively.

The prevalence of cytogenetic abnormalities, including CK, is not homogeneously spread across all TP53 ^mut MN. Instead, there is a complex interplay between CK, blast percentage, VAF, and allelic status (Figure 3C, Shah et al., Blood Adv, in print) [26, 57]. Given this complex interaction, a hierarchical prognostic model that simultaneously accounts for multiple factors is needed.

• 4Complex karyotype is a practical surrogate of multi‐hit TP53 inactivation. In the absence of comprehensive CNV analysis, CK (in the context of TP53 ^mut VAF ≥ 10%) is considered a multi‐hit equivalent by ICC, but not by WHO‐5. We have reported that the survival of MDS cases with one TP53 ^mut VAF < 50% with CK was comparable to those with 17p13.1 deletion on karyotype (10.4 vs. 11.0 months; p = 0.39) but was significantly poorer than those cases with monoallelic TP53 ^mut without 17p loss or CK (10.4 vs. 33.4 months; p < 0.0001), indicating that for MDS cases with single TP53 ^mut VAF < 50%, the presence of CK can be considered a practical surrogate for biallelic TP53 inactivation.
• 5Verification of 17p13.1 deletion detected by metaphase karyotype by copy number variation analysis. WHO‐5 mandates verification of 17p13.1 deletion detected by metaphase karyotype by an additional CNV method. In MDS with a single TP53 ^mut with VAF < 50%, mere detection of 17p13.1 del on karyotype is not considered to be evidence of biallelic inactivation, while ICC does not mandate such verification.

Verification of loss across the TP53 locus by CNV analysis is useful, especially in cases with complex structural rearrangements of 17p13.1 on karyotype. Second, metaphase karyotype can miss cryptic LOH/cnLOH.

Given the wide spectrum of instruments and personnel availability at different labs, an algorithmic approach that is highly dependent upon resources, turn‐around time, and institutional practices, timely CNV analysis is fraught with practical limitations: the majority of diagnostic laboratories cannot reliably analyze CNV using the routinely used NGS. SNP array and/or FISH studies would be pursued after TP53 ^mut status is known. These sequential tests can substantially delay the risk stratification and management decisions.

This was illustrated in a recent international study of 603 MN harboring TP53 ^mut MN, where ~20% had CNV analysis (Shah et al., under review). Of the evaluable cases with 17p13.1 deletion on karyotype, LOH across the TP53 locus could be confirmed by CNV analysis in 94% of cases. Moreover, in cases without 17p13.1 deletion on karyotype, CNV analysis detected LOH/cnLOH across the TP53 locus in 22.5% of cases. Importantly, all these cases harbored CK, and none of the cases without CK and 17p13.1 deletion on metaphase karyotype had LOH/cnLOH across the TP53 locus. Furthermore, the survival of single TP53 ^mut MDS with VAF < 50% and 17p13.1 del on metaphase karyotype was similar to cases with biallelic inactivation. Collectively, these results indicated that in clinical practice, 17p13.1 deletion on metaphase cytogenetics can be considered as evidence of the loss of the trans allele.

• 6Morphological assessment. WHO‐5 requires that a case meets the MDS diagnostic criteria before it can be further classified as MDS with biallelic TP53 inactivation (MDS‐biTP53) [98]. On the other hand, ICC guidelines allow for the classification of multihit TP53 ^mut cytopenic cases harboring < 5% blasts in the absence of dysplastic features as MDS with mutated TP53. This recommendation was based on a study that demonstrated that CCUS cases harboring TP53 ^mut that lack sufficient dysplasia to diagnose MDS share biologic features, mutational patterns, and survival with TP53 ^mut MDS [102]. Recent larger, population‐based studies did not confirm TP53 ^mut as an independent adverse risk factor for survival [103, 104]. Finally, in a single institutional study of 29 TP53 ^mut CCUS patients, only 3 (10.3%) progressed to MN, and the progression‐free and overall survival of TP53 ^mut CCUS were comparable to TP53 ^wt CCUS at up to 5 years [105]. Therefore, additional data is required to support future recommendations.
• 7Impact of prior cytotoxic therapies. The two classifications differ in the interpretation of whether prior cytotoxic therapy modulates the risk‐stratification. ICC removed the subcategory of “therapy‐related myeloid neoplasms,” substituting it with diagnostic qualifiers instead [86]. The WHO‐5 has grouped t‐MN with secondary MN and renamed it as “myeloid neoplasm post‐cytotoxic therapy” [98]. In either cases, no distinctions were made between de novo and post‐cytotoxic therapy TP53 ^mut MN. These recommendations were based on an international study of 229 t‐MDS, of which 18% (n = 41) harbored TP53 ^mut that showed inferior survival of multi‐hit TP53 ^mut t‐MDS compared to single‐hit TP53 ^mut [24]. In contrast, analysis of a larger TP53 ^mut t‐MN (n = 260, 34.2% TP53 ^mut) showed that unlike in de novo TP53 ^mut MN, clinical features, structural chromosomal abnormalities, as well as co‐mutation pattern were comparable between single‐ and multi‐hit TP53 ^mut t‐MN [9]. Critically, unlike de novo MDS [24], the incidence of AML progression and survival of single‐hit TP53 ^mut t‐MDS were comparable to multi‐hit TP53 ^mut t‐MDS.

These findings challenge the underlying assumption of the ICC classification that TP53 ^mut MN—regardless of the underlying etiology—has similar genomic characteristics and outcomes and cautions against the underestimation of the poor prognosis of single‐hit TP53 ^mut t‐MDS [9].

Based on the evidence presented above [9, 26, 100], we propose a hierarchical model (Shah et al., Blood Adv, in print). Since TP53 ^mut BM/PB ≥ 5% with VAF < 10% and no CK are relatively uncommon (4.2%), all TP53 ^mut MN (VAF ≥ 2%) can be considered TP53 ^mut MN. Accepting this stipulation, a simplified and clinically useful model is proposed (Figure 4) that acknowledges poor survival of 91.9% TP53 ^mut MN.

FIGURE 4.

Evidence‐based hierarchical classification of TP53‐mutated (TP53 ^mut) myeloid neoplasms (MN). AML—acute myeloid leukemia; CK—complex karyotype; EB—excess blasts; LB—low (< 5% blasts); MDS—myelodysplastic syndrome; OS—overall survival (calculated from diagnosis); VAF—variant allele frequency. Adapted from Shah et al. (Blood Adv, in print).

6. Management of TP53‐Mutated Myeloid Neoplasms

Effective management of TP53 ^mut MN continues to remain a long‐standing challenge, primarily due to its intrinsic chemorefractory nature [8, 9, 10]. Thus, despite its prevalence in cancer, TP53 remains undruggable. The notable progress made in other forms of MN has not benefited TP53 ^mut MN and survival following the diagnosis remains < 1 year. Commonly used strategies include HMA, intensive chemotherapies, and BCL‐2 inhibitor venetoclax. Initial studies of venetoclax—given high response rates—raised the hope that it would improve outcomes of TP53 ^mut AML. However, with longer follow‐up, it is clear that venetoclax‐based regimens do not lead to a meaningful improvement in the survival of TP53 ^mut AML [11, 12].

The lack of progress was evident in a recent meta‐analysis that observed comparable survival (6.5, 6.1, and 6.2 months) with intensive chemotherapy, HMA, and venetoclax plus HMA, respectively [106]. Disconcertingly, further development of several classes of drugs has been terminated due to lack of efficacy [13]. Below we summarize the published experience of commonly used management strategies.

1. Supportive care. Given age, comorbidities, concurrent cytopenia, and enrichment of t‐MN, 9.4%–21% of patients do not pursue disease‐modifying therapies, resorting to supportive care alone [9, 46]. Outcomes with a palliative care approach are poor as expected, with a median survival of 1–4 months [9, 46].

2. Intensive chemotherapy. Response to cytotoxic chemotherapies is highly dependent on the presence of intact p53 to enable the induction of apoptosis [107, 108]. Hence, TP53 mutated MN respond poorly to chemotherapy. Historically, the combination of anthracyclines with cytarabine or a high‐dose cytarabine‐based regimens was used as the frontline therapy for patients deemed fit to receive intensive therapy [109]. The uptake of intensive chemotherapy is traditionally lower in TP53 ^mut MN (14.6%–22.5%) [9, 46, 57], reflecting both patient and disease‐related factors: age, frailty, aggressiveness, and inherent chemorefractoriness that is characteristic of TP53 ^mut. Model response rate (20%–40%) and median survival (5–11 months) have been reported by multiple groups [8, 9, 57]. Interestingly, the presence of hotspot TP53 ^mut was associated with an inferior response rate compared to those with non‐hotspot mutations (17.9% vs. 57.1%; p = 0.025) [46].

In a single‐institution study of 202 TP53 ^mut AML, 22% received intermediate/high‐dose cytarabine‐based regimens. The response rate and survival in this cohort correlated with TP53 ^mut VAF: those with VAF ≤ 40% had a median survival of 18.1 months compared to 5 months for patients with VAF > 40%. Moreover, the benefit of receiving a higher‐intensity therapy (compared to HMA) was limited to patients with VAF ≤ 40% [110]. Whether these findings represent a true therapeutic benefit of cytarabine‐based regimens or a representation of a younger cohort who underwent allogeneic transplant is unclear. Therefore, independent validation of this observation is awaited.

CPX‐351 is an encapsulated formulation of liposomal daunorubicin and cytarabine that preferentially delivers a synergistic 5:1 drug ratio into the leukemia cells, minimizing off‐target toxicities to the normal bone marrow cells. CPX‐351 is approved by the FDA for secondary AML and t‐AML—both of which are enriched for TP53 ^mut, though stratification based on TP53 ^mut was not performed [111]. As with other modalities, composite complete responses were half in TP53 ^mut compared to TP53 ^wt AML (33% vs. 66%, p = 0.035) [112].

• 3HMA. HMA such as 5‐azacitidine and decitabine represent one of the most prevalent strategies for MN including TP53 ^mut MN. HMA were considered the preferred frontline therapy in elderly/unfit patients before the venetoclax era [113]. A retrospective analysis of the ASTRAL‐1 trial that included 17% TP53 ^mut AML identified TP53 ^mut as an “adverse‐risk feature” [114]. The poor outcome was confirmed in an independent retrospective analysis [115]. Welch et al. reported that 10‐day decitabine therapy ameliorated the adverse impact of TP53 ^mut resulting in comparable survival of TP53 ^mut and TP53 ^wt (12.7 vs. 15.4 months, p = 0.79) [116]. However, a similar benefit could not be validated in another study that reported comparable response rate (40% vs. 43%, p = 0.78) and survival (6 vs. 5·5 months) with 10‐ vs. 5‐day decitabine regimen [117].
• 4Venetoclax. Venetoclax, an orally available selective BCL‐2 inhibitor, has revolutionized the treatment for elderly AML patients or those ineligible for intensive induction. Pivotal studies VIALE‐A [118] and VIALE‐C [119] that studied venetoclax in combination with 5‐azacitidine and low‐dose cytarabine (LDAC), respectively, showed higher composite response rates (CR + CRi) with the addition of venetoclax (55.3% vs. 0% in VIALE‐A [118] and 18% vs. 0% in VIALE‐C) [119]. However, it was soon realized that decreased expression of BAX in TP53‐deficient AML cells contributed to their inherent resistance to BH3 mimetics [120]. In addition, monocytic differentiation and downregulation of HLA‐DR and CD34 were suggested as a potential mechanisms of acquired resistance [121].

Upon longer follow‐up, it is now quite clear that venetoclax‐based regimens do not improve outcomes for this subset [122, 123, 124]. In a recent study of 301 newly diagnosed AML patients treated with venetoclax plus HMA, multi‐hit TP53 ^mut was associated with lower CR/CRi compared to single‐hit TP53 ^mut or TP53 ^wt AML (38% vs. 63% vs. 67%). TP53 ^mut also conferred an inferior relapse‐free survival (7.9 vs. 19.3 months) and overall survival (5.9 vs. 16.6 months) compared to TP53 ^wt AML [125]. Another pooled analysis of 279 patients treated on the pivotal venetoclax plus azacitidine studies confirmed TP53 ^mut as an independent risk factor for survival (median 5.5 months). Combined, these observations have led to TP53 ^mut as the defining feature of the “adverse‐risk” category in the ELN risk classification for patients receiving less‐intensive therapies [126].

Given the FDA approval, the majority of the evidence for the efficacy of venetoclax‐based regimen is limited to AML. Recently, venetoclax combinations have been tried in high‐risk MDS including TP53 ^mut MDS [127]. In a recent phase 1b study of venetoclax with azacitidine for treatment‐naive HR‐MDS (NCT02942290), 20 (18.7%) of 107 patients harbored TP53 ^mut. TP53 ^mut cases had a comparable rate of complete remission to the entire cohort (25% vs. 29%), though survival was numerically inferior (11.2 vs. 26 months). Longer follow‐up and larger studies will be needed to ascertain these findings.

• 5Allogeneic hematopoietic cell transplant. AlloHCT is the only modality with curative potential for patients with TP53 ^mut MN. A retrospective analysis of the Blood and Marrow Transplant Clinical Trials Network 1102 study that randomized high‐risk MDS patients, including TP53 ^mut MDS, based on donor availability showed significantly improved long‐term survival in patients undergoing alloHCT compared to non‐transplant approaches (3‐year OS 23% vs. 11%, p = 0.04). Therefore, alloHCT remains the “gold‐standard” for all eligible patients [128].

On the other hand, given the resource‐intense nature as well as high morbidity and mortality, there is an active debate around the applicability of TP53 ^mut MN [129, 130]. Historically, only 7%–18% of TP53 ^mut MN undergo alloHCT (Figure 5A) [9, 46, 111, 131]. In the absence of systematic studies, the reasons for the low utilization are speculative and include older age, frailty, comorbidities, concurrent malignancies, inability to achieve the desired pre‐alloHCT response, and high post‐alloHCT relapse risk.

FIGURE 5.

Outcomes following allogeneic hematopoietic cell transplant (alloHCT) for myeloid neoplasms harboring TP53 mutations (TP53 ^mut). (A) Utilization of alloHCT remains low in TP53 ^mut MN. (B) Three‐year overall survival following alloHCT for TP53 ^mut MN across published studies. Studies include: Badar et al. [131], Baranwal et al. (in print), Bernard et al. [24], Grob et al. [8], Ibrahim et al. [132], Kaur et al. [46], Lindsley et al. [2], Loke et al. [133], Lontos et al.** [134], Middeke et al. [135], Shah et al. [57], and Yoshizato et al. [136] 17p del—17p deletion; CK—complex karyotype; MH—multi‐hit; SH—single‐hit. **Overall survival at 2 years.

Three‐year survival is 10%–15% in most published studies regardless of the MDS or AML phenotype (Figure 5B). A recent international multicenter study (7 transplant centers in USA and Australia, Baranwal et al., in print) of 134 TP53 ^mut MN patients who underwent alloHCT confirmed poor survival in the contemporary era: median post‐HCT survival was 1.03 years and OS at 1‐, 2‐, and 3‐years was 51.4%, 35.1%, and 25.1%, respectively. Multihit TP53 inactivation was associated with significantly shorter 3‐year survival compared to those with non‐multihit TP53 inactivation (16.9% vs. 54.9%, p = 0.002). Interestingly, non‐DBD TP53 ^mut only and DNMT3A co‐mutation were associated with 3.4‐ and 2.6‐fold lower relapse‐free survival (RFS). These results, for the first time, suggested the mutational and co‐mutational patterns associated with post‐HCT outcomes. These results require validation in an independent cohort.

Similar findings are reported in a single‐center analysis of 240 MN cohort [134]. Survival was comparable between TP53 ^mut MDS and AML, and 2‐year OS for the entire cohort was 34%. Hierarchical analysis identified 3 groups: best survival was in cases with TP53 ^mut VAF < 50% that did not harbor CK/deletion 5q/7q, followed by TP53 ^mut VAF < 50% that harbored CK/deletion 5q/7q, and the worst being the cases with TP53 ^mut VAF ≥ 50% with CK/deletion 5q/7q (2‐year OS of 60%, 22%, and 3% respectively).

Post‐transplant mortality and morbidity are driven by the high relapse rate. Therefore, multiple studies have attempted to identify factors associated with lower post‐alloHCT relapse. Unfortunately, no clear picture emerges, though the younger recipient age [2], TP53 ^mut VAF < 40% [110], single‐hit TP53 ^mut (Baranwal et al., in print), absence of CK [133], undergoing alloHCT in 1st remission [131], achieving CR at day +100 [46, 131], and the development of moderate to severe GVHD [46, 131] were associated with favorable survival. Therefore, there is an active debate if (a) the disease status at alloHCT and (b) the choice of conditioning intensity impact post‐alloHCT survival.

As to the former, in TP53 ^mut EB‐2/AML patients treated with intensive chemotherapy, achieving minimal residual disease (MRD) negativity was not associated with longer post‐alloHCT survival [8]. In another multicenter study, achieving MRD negative CR before alloHCT was not associated with improved relapse‐free or overall survival [131]. On the other hand, in a study by Hunter et al. [137], of the 16 TP53 ^mut MN who underwent alloHCT, 7 achieved a clearance of TP53 ^mut (by NGS) pre‐alloHCT and had a numerical trend towards improved survival (25.2 vs. 11.7 months). Interestingly, patients with pre‐alloHCT TP53 ^mut clearance benefited from alloHCT over continued HMA (25.2 vs. 7.7 months); whereas those with clonal persistence had survival comparable between the two approaches (11.7 vs. 7.7 months). If validated, these studies can be the first step in identifying suitable candidates.

Second, the choice of optimal conditioning is far from certain: while some studies suggested a reduced risk of relapse using myeloablative conditioning (MAC), most do not demonstrate a survival benefit [2, 131, 138]. For example, in a CIBMTR study of TP53 ^mut MDS, both the risk of relapse and survival were comparable between MAC and reduced‐intensity conditioning (RIC) [2]. This was recently confirmed by another multi‐institutional study [131]. In contrast, others have shown a reduced risk of relapse but a higher risk of non‐relapse mortality with MAC, culminating still in a survival comparable to RIC regimens [138, 139].

Finally, two studies suggest an advantage of melphalan‐based conditioning. In a single‐institution study of t‐MDS patients undergoing alloHCT predominantly using melphalan‐based conditioning showed higher RFS and OS were shown compared to other studies, though the impact of melphalan could not be isolated [132]. In the multicenter study discussed above, the inclusion of melphalan was associated with improved RFS (HR 0.52, p = 0.005). The discrepancies above may—at least in part—be explained as the benefit of melphalan‐based conditioning (almost exclusively used in the RIC context) was limited to cases with < 5% blasts pre‐alloHCT (Baranwal et al., Blood Adv, in print).

6.1. Immunotherapies and Novel Targets

6.1.1. Immunotherapy

Immunotherapy has taken center stage in oncology over the last decade. Immunotherapy agents rely on the innate ability of the immune system to detect and eliminate tumor cells. Checkpoint inhibitors function by blocking inhibitory co‐receptors on T‐cells, including programmed cell death (PD)‐1, programmed cell death ligand (PD‐L)‐1, and cytotoxic T‐lymphocyte antigen (CTLA)‐4 inhibitors. These proteins are accessories to immune evasion by tumor cells, and the expression of these receptors and their ligands is augmented in myeloid malignancies, suggesting a possible resistance mechanism to conventional therapies [140, 141]. As discussed above, TP53 ^mut MN have a distinct immune milieu—raising the possibility that immune‐based therapies can be exploited in the treatment of TP53‐mutated AML. Current approaches under study include ICI, bispecific and dual‐antigen–receptor targeting antibodies, chimeric antigen receptor (CAR) T‐cell therapy, and newer targets such as T‐cell immunoglobulin and mucin domain (TIM)‐3 inhibitors. The stimulator of interferon genes are some novel treatment modalities under this domain [142].

1. ICI. Contrary to solid malignancies and some hematological malignancies, TP53 ^mut MN are modestly sensitive to ICI when combined with HMA or intensive chemotherapy. In a phase 2 study, nivolumab was used in combination with 5‐azacitidine in relapsed/refractory AML, 16 of whom harbored TP53 ^mut. The overall response rate for TP53 ^mut AML was 19% (3/16 patients) [41]. Similarly, another phase II trial showed that pembrolizumab after high‐dose cytarabine was safe in 37 patients with R/R AML (overall response rate 46%, complete remission 38%, median overall survival 11.1 months). Two of the 5 (40%) patients with TP53 ^mut in this cohort achieved complete remission [42]. Then, in a phase II trial of upfront nivolumab with idarubicin and cytarabine in 44 treatment‐naïve patients with high‐risk MDS/AML, including 8 with TP53 ^mut MDS, the CR/CRi rate was 78% [43] Prospective, randomized studies are needed to assess if patients with TP53 ^mut AML have better outcomes with ICI combinations than with conventional chemotherapies alone. Designing these trials will need to take a broader view as most patients will be planned to undergo alloHCT, potentially raising the concern for exacerbating graft‐vs.‐host disease [143].

2. Sabatolimab is a monoclonal antibody targeting TIM‐3. In a phase Ib, multi‐arm, open‐label, multicenter study for AML, high/very high‐risk MDS, and chronic myelomonocytic leukemia, 14 evaluable patients with TP53 ^mut, the overall response rate was 71.4%, complete remission was seen in 4 (28.6%) and the median duration of response was 21.5 months [144]. STIMULUS‐MDS‐1 [145] was a multicenter, randomized, double‐blind, placebo‐controlled, phase II study in newly diagnosed high‐risk MDS including TP53 ^mut MDS. Of 186 patients screened, 36% of the sabatolimab and 33% of the placebo cohort harbored TP53 ^mut. While response rates or survival were not stratified by TP53 ^mut status, the primary endpoints were not met. Complete remission rate (22% vs. 18%, p = 0.77) and progression‐free survival (11.1 vs. 8.5 months, p = 0.1) were comparable in the sabatolimab and placebo groups. Ultimately, the phase III study (STIMULUS‐MDS‐2) failed to meet the primary endpoint of improved overall survival—resulting in the sponsor announcing discontinuation of the sabatolimab program [45].

3. Cluster of differentiation (CD)‐47 is widely expressed across various cell types that interacts with signal‐regulatory protein (SIRP)‐α on phagocytic cells, delivering an anti‐phagocytic “don't‐eat‐me” signal. Its overexpression in cancer cells is hypothesized to counteract the pro‐phagocytic signals, avoiding phagocytosis. Magrolimab, a humanized immunoglobulin G4 anti‐CD47 monoclonal antibody, inhibits the CD47‐SIRPα interaction, enhancing cancer cell phagocytosis. In combination with 5‐azacitidine that upregulates calreticulin on AML cells, increasing the “eat me” signal. Based on the in vitro and in vivo preclinical data, a phase I study of magrolimab monotherapy was well tolerated in relapsed/refractory AML (NCT02678338, CAMELLIA study) [146] demonstrated satisfactory safety. Phase Ib (5F9005, NCT03248479) study combining magrolimab with azacitidine in newly diagnosed AML who were ineligible for intensive chemotherapy. Of 87 patients, 82.8% harbored TP53 ^mut (median VAF 61%, range: 9.8–98.7). After a median of 4 (range, 1–39) cycles of therapy, complete remission rate was 31.9% in TP53 ^mut AML. The median overall survival of TP53 ^mut and TP53 ^wt AML were 9.8 and 18.9 months, respectively. The median duration of response was 7.7 months. Among 14 TP53 ^mut patients who achieved measurable residual disease (MRD)‐negative responses, median survival was 14.5 months compared to 7.5 months who remained MRD‐positive [147]. Subsequent randomized study of TP53 ^mut AML patients who were ineligible for intensive chemotherapy were treated with magrolimab plus AZA vs. investigator choice (ENHANCE‐2) and AZA+ VEN+ magrolimab vs. AZA+VEN [148]. However, magrolimab trials have been discontinued due to futility based on planned analysis [149]. Similarly, trials using other anti‐CD47 monoclonal antibody (evorpacept, NCT04417517 and NCT04755244) have been terminated as the combination “did not substantially improve upon the historical activity of azacitidine alone.” Currently, there is one open trial of anti‐CD47 antibody with HMA in AML and high‐risk MDS (NCT06008405, Table 4), demonstrating continued interest in this target.

4. Bispecific antibodies, bispecific T‐cell engagers, and dual affinity retargeting antibodies have been under exploration for the management of TP53 ^mut AML. Flotetuzumab, a CD123xCD3‐targeting dual affinity retargeting antibody that works by enhancing the formation of an immunologic synapse between cytotoxic T‐ and AML cells independent of the major histocompatibility complex (MHC) pathway, has shown promising efficacy in TP53 ^mut (complete remission 47%, median survival 10.3 months) [150]. Another bispecific antibody against CD123xCD3 (APVO436‐5001) has also shown efficacy and safety in an early‐phase clinical trial [151].

5. Chimeric antigen T‐cell receptor therapy: Compared to other hematological malignancies, the development of CAR‐T therapy in MN has been confronted by toxicities including myelosuppression and poor CAR T‐cell persistence [152]. Despite that, multiple trials targeting established (CD33, CD123) and novel (CLL‐1, CD371) and combination targets are being investigated (Table 4).

TABLE 4.

Select ongoing clinical trials for TP53‐mutated myeloid neoplasms.

NCT identifier	Title	Status	Condition	Intervention(s)	Phase
NCT06778187	Oral‐ATO for TP53‐mutated Myeloid Malignancies	Not yet Recruiting	AML, MDS	Oral arsenic trioxide	2
NCT04277442	Testing Nivolumab in Combination with Decitabine and Venetoclax in Patients with Newly Diagnosed TP53 Gene Mutated Acute Myeloid Leukemia	Active, not Recruiting	AML	Decitabine, nivolumab, venetoclax	1
NCT06456463	A Study of Tagraxofusp in Combination with Venetoclax and Azacitidine in Adults with Untreated CD123+ Acute Myeloid Leukemia Who Cannot Undergo Intensive Chemotherapy	Recruiting	AML	Tagraxofusp, venetoclax, 5‐azacitidine	2
NCT05396859	Entrectinib in Combination with ASTX727 for the Treatment of Relapsed/Refractory TP53 Mutated Acute Myeloid Leukemia	Recruiting	AML	Decitabine, cedazuridine, entrectinib	1
NCT06549790	Study of NMS‐03597812 in Adult Patients with Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	NMS‐03597812	1
NCT03772925	Pevonedistat and Belinostat in Treating Patients with Relapsed or Refractory Acute Myeloid Leukemia or Myelodysplastic Syndrome	Active, not Recruiting	AML, MDS	Belinostat, pevonedistat	1
NCT02665065	Study of Iomab‐B vs. Conventional Care in Older Subjects with Active, Relapsed or Refractory Acute Myeloid Leukemia	Active, not Recruiting	AML	Iomab‐B	3
NCT04358393	A Study of APG‐115 Alone or Combined with Azacitidine in Patients With AML, CMML, or MDS	Recruiting	AML, MDS	APG‐115, 5‐azacitidine	1/2
NCT06514261	Testing the Combination of an Anti‐Cancer Drug, Iadademstat, With Other Anti‐Cancer Drugs (Venetoclax and Azacitidine) for Treating AML	Recruiting	AML	Azacitidine, iadademstat, venetoclax	1
NCT04477291	A Study of CG‐806 in Patients with Relapsed or Refractory AML or Higher‐Risk MDS	Active, not Recruiting	AML, MDS	CG‐806	1
NCT03560882	A Pilot Trial of Atorvastatin in Tumor Protein 53 (p53)—Mutant and p53 Wild‐Type Malignancies	Active, not Recruiting	AML, MDS	Atorvastatin	1
NCT03850574	Clinical Trial to Evaluate the Safety, Tolerability, Pharmacokinetics and Pharmacodynamics of Tuspetinib (HM43239) in Patients with Relapsed or Refractory Acute Myeloid Leukemia	Recruiting	AML, MDS	Tuspetinib, venetoclax, 5‐azacitidine	1/2
NCT06008405	Clinical Trial Evaluating the Safety of the TQB2928 Injection Combination Therapy	Recruiting	AML, MDS	TQB2928 (anti‐CD47)	1
NCT06130579	Interferon‐alpha for TP53 Myeloid Malignancy Post Allo‐HSCT	Recruiting	AML, MDS	IFN‐alpha	2
NCT04219163	Chimeric Antigen Receptor T‐cells for the Treatment of AML Expressing CLL‐1 Antigen	Recruiting	AML	CLL‐1 CAR	1
NCT05672147	CD33‐CAR T Cell Therapy for the Treatment of Recurrent or Refractory Acute Myeloid Leukemia	Recruiting	AML	CD33 CAR	1
NCT04923919	Clinical Study of Chimeric Antigen Receptor T Lymphocytes (CAR‐T) in the Treatment of Myeloid Leukemia	Recruiting	AML	CLL‐1 CAR	1
NCT06326021	Optimised CD33 (FL‐33) CAR T Therapy for Refractory/Relapsed Acute Myeloid Leukaemia	Recruiting	AML	FL‐33 CAR	1
NCT05017883	TAA05 Cell Injection in the Treatment of Recurrent/Refractory Acute Myeloid Leukemia	Recruiting	AML	TAA05 CAR
NCT05984199	Donor‐Derived Anti‐CD33 CAR T Cell Therapy (VCAR33) in Patients with Relapsed or Refractory AML After Allogeneic Hematopoietic Cell Transplant	Recruiting	AML	VCAR33	1/2
NCT06765876	CART123 T Cells in Relapsed or Refractory CD123+ Hematologic Malignancies: A Dose Escalation Phase I Trial	Recruiting	AML	CAR123, autologous	1
NCT06197672	Chimeric Antigen Receptor T Cell Redirected to Target CD4 Positive Relapsed Refractory Acute Myeloid Leukemia (AML) As a Bridge to Allogeneic Stem Cell Transplant	Recruiting	AML	CD4, autologous	1
NCT05457010	Phase I Study of Cell Therapies for the Treatment of Patients with Relapsed or Refractory AML or High‐risk MDS	Recruiting	AML	SPRX002/ARC‐T cells	1
NCT06125652	Administration of Anti Tim‐3/CD123 CAR‐T Cell Therapy in Relapsed and Refractory Acute Myeloid Leukemia (rr/AML)	Recruiting	AML	TIM‐3/CD123 CAR	1/2
NCT03971799	Study of Anti‐CD33 Chimeric Antigen Receptor‐Expressing T Cells (CD33CART) in Children and Young Adults with Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CD33 autologous/allogeneic CAR	1/2
NCT05488132	Administration of Anti‐siglec‐6 CAR‐T Cell Therapy in Relapsed and Refractory Acute Myeloid Leukemia (rr/AML)	Recruiting	AML	siglec‐6 CAR	1/2
NCT06609928	FH‐FOLR1 Chimeric Antigen Receptor T Cell Therapy for Treating Pediatric Patients with Relapsed or Refractory Acute Myeloid Leukemia	Recruiting	AML	FOLR1 CAR	1
NCT05949125	Phase 1 Study of Allo‐RevCAR01‐T‐CD123 in Patients with Selected CD123 Positive Hematologic Malignancies	Recruiting	AML	CD123, allogeneic CAR	1
NCT04230265	Phase 1 Study of UniCAR02‐T‐CD123 in Patients with Selected CD123 Positive Hematologic Malignancies	Recruiting	AML	CD123 CAR	1
NCT04265963	CD123‐Targeted CAR‐T Cell Therapy for Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CD123 CAR	1/2
NCT06762132	A Clinical Study to Explore the Safety and Efficacy of CD33 CAR‐T Cell in Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CD33 CAR	1
NCT04272125	Safety and Efficacy of CD123‐Targeted CAR‐T Therapy for Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CD123 CAR	1/2
NCT04803929	Clinical Study of Anti‐ILT3 CAR‐T Therapy for R/R AML(M4/M5)	Recruiting	AML	ILT3 CAR	1
NCT03190278	Study Evaluating Safety and Efficacy of UCART123v1.2 in Patients with Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	UCART123v1.2	1
NCT06017258	A Study of CD371‐YSNVZIL‐18 CAR T Cells in People with Acute Myeloid Leukemia	Recruiting	AML	CD371‐specific/YSNVz/I‐18 CAR	1
NCT05748197	A Study of ADCLEC.syn1 in People with Acute Myeloid Leukemia	Recruiting	AML	ADCLEC.syn1	1
NCT06642025	EX02 CAR‐T Cells for Relapsed and Refractory Acute Myeloid Leukemia	Recruiting	AML	EX02 CAR	1
NCT06709131	A Clinical Study to Explore the Safety and Efficacy of CT0991 in Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CT0991 CAR	1
NCT06420063	Sequential CAR‐T Cells Targeting CD33/CD123 in Patients with Acute Myelocytic Leukemia AML	Recruiting	AML	CD33/CD123 CAR	1/2
NCT03291444	CAR‐T Cells Combined with Peptide Specific Dendritic Cell in Relapsed/Refractory Leukemia/MDS	Recruiting	AML	CAR	1
NCT05995041	Universal CAR‐T Cells Targeting AML	Recruiting	AML	CLL‐1, CD33, CD38 and/or CD123 CAR	1
NCT05945849	CD33KO‐HSPC Infusion Followed by CART‐33 Infusion(s) for Refractory/Relapsed AML	Recruiting	AML	CD33 CAR	1
NCT06128044	CRISPR‐Edited Allogeneic Anti‐CLL‐1 CAR‐T Cell Therapy in Patients with Relapsed/Refractory Acute Myeloid Leukemia	Recruiting	AML	CLL‐1 CAR	1
NCT04662294	CD 70 CAR T for Patients with CD70 Positive Malignant Hematologic Diseases	Recruiting	AML	CD70 CAR	1
NCT05105152	PLAT‐08: a Study of SC‐DARIC33 CAR T Cells in Pediatric and Young Adults with Relapsed or Refractory CD33+ AML	Recruiting	AML	CD33 CAR	1
NCT04318678	CD123‐Directed Autologous T‐Cell Therapy for Acute Myelogenous Leukemia (CATCHAML)	Recruiting	AML	CD123 CAR	1
NCT05377827	Dose‐Escalation and Dose‐Expansion Study to Evaluate the Safety and Tolerability of Anti‐CD7 Allogeneic CAR T‐Cells (WU‐CART‐007) in Patients with CD7+ Hematologic Malignancies	Recruiting	AML	CD7 CAR	1

Note: List retrieved from clinicaltrials.gov on February 22, 2025.

Abbreviations: AML—acute myeloid leukemia; CAR—chimeric antigen T‐cell receptor therapy; MDS—myelodysplastic syndrome.

6.1.2. Novel Targeted Therapies

1. TP53‐activator: Eprenetapopt (APR‐246) is a novel, first‐in‐class, small molecule that restores wild‐type p53 functions in TP53‐mutant cells. It was evaluated in a phase Ib/II study. Of 55 patients (40 MDS, 11 AML, 4 MDS/MPN) with at least one TP53 mutation^mut treated, the overall response rate was 71%, with 44% achieving CR. Among MDS patients, 50% achieved CR, and 58% had a cytogenetic response. The overall response rate and complete remission rate in AML were 64% and 36%, respectively. Interestingly, responders had a significant reduction in TP53 ^mut VAF, with 38% achieving complete molecular remission. Median overall survival was 10.8 months. Thirty‐five percent of patients underwent allogeneic hematopoietic stem‐cell transplant, with a median overall survival of 14.7 months [153]. Based on these exciting results, a combination phase I multicenter study with HMA and venetoclax was conducted. Forty‐nine patients were enrolled. The combination was well tolerated, and the overall response with eprenetapopt and venetoclax with azacytidine was 38% [154]. Despite these encouraging results, further development of the drug is unfortunately halted due to futility following a phase III study [155].

2. Splenic tyrosine kinase inhibitors: Splenic tyrosine kinase (Syk) is overexpressed in AML, and the overactivity is associated with poor prognosis. Entospletinib is a Syk inhibitor that was evaluated in combination with decitabine for TP53 ^mut or CK AML in the BEAT AML Study. The drug combination showed modest activity—leading to discontinuation due to futility [156].

3. ROS1 inhibitor: Various malignancies, including AML, demonstrate aberrant expression of the ROS proto‐oncogene 1 receptor tyrosine kinase, making it a viable target for anticancer therapies. TP53‐deficient cells show sustained sensitivity to the ROS1 inhibitor entrecitinib [157]. Based on these results, a phase I study is evaluating entrecitinib, in combination with a hypomethylating agent, as a treatment for relapsed/refractory TP53 ^mut AML (Table 4).

4. Nutlin analogs: Given the central role of MDM2 as a negative regulator of TP53, small molecular inhibitors of MDM2, such as nutlin analogs, have been explored for TP53 ^mut MN. In the Phase 1b trial of idasanutlin plus venetoclax in relapsed/refractory AML, an overall response rate was seen in 3 of 10 TP53 ^mut (no complete remissions), and the median duration of response and survival were merely 2.3 and 3.67 months. TP53 ^mut was subclonal in responders. At discontinuation, 25 TP53 ^mut were noted in 12 patients—of which 22 were pre‐existing [158]. The phase 3 MIRROS trial was a multicenter, randomized, double‐blind, phase 3 study of the MDM2 antagonist idasanutlin plus cytarabine in relapsed/refractory AML. The study enrolled patients regardless of TP53 ^mut status with the rationale that some TP53 ^mut could retain wild‐type TP53 function. Approximately 15% of patients enrolled harbored TP53 ^mut. Endpoints, including overall survival (median, 8.3 vs. 9.1 months p = 0.58), complete remission (20.3% vs. 17.1%), and overall response rate (38.8% vs. 22.0%) were all comparable between idasanutlin plus cytarabine compared to placebo plus cytarabine [159]. Overall, consistent with the mechanism, these results suggest the possibility that MDM2 inhibitors may not be effective in TP53 ^mut MN.

The above encounters highlight the unparallelled therapeutic challenges presented by TP53 ^mut MN. We conclude that the development of effective therapies for TP53 ^mut MN is an urgent unmet clinical need.

7. Future Directions

An unwavering interest of scientists and clinicians alike in all aspects of TP53 ^mut MN is a testament to the enormity of the challenge. The review above highlights advances made in our understanding of the disease, but also substantial challenges that persist. It is hoped that the recognition as a separate entity will stimulate research, facilitate clinical trial enrollment, and fuel drug discovery. To that end, we propose the following as the areas of the highest priority.

Not all TP53 ^mut CH progress to MN, and the typical latency seen is > 5 years. Documented stability of the TP53 ^mut clone without leukemic transformation and the dramatic contrast between the outcomes of TP53 ^mut CH/CCUS and TP53 ^mut MN, collectively highlight the critical need to characterize the factors associated with leukemic transformation. Specifically, the availability of biomarker(s) will help identify high‐risk patients and will guide surveillance strategy. This is even more critical for patients with known TP53 ^mut CH in need of informed decision to undergo cytotoxic therapy for unrelated malignancy. Ultimately, it is hoped that early interventions such as pre‐emptive alloHCT before the acquisition of biallelic TP53 inactivation and leukemic transformation would improve outcomes.

Second, future iterations of the TP53 ^mut MN classification are expected to incorporate the emerging evidence. The current diagnostic criteria aim to overcome the technical limitation of assigning the biallelic loss confidently, necessitating the use of surrogates such as “presumed biallelic” or “multihit equivalent.” In the future, incorporation of WES/WGS or preferentially single‐cell studies will help localize the cell‐of‐origin and localize biallelic TP53 alterations in leukemic cells. While such technology is available, a globally adopted classification must balance feasibility (availability, personnel need, turn‐around‐time, and cost) with accuracy.

Third, with near‐universal use of NGS and progressively increasing sensitivity, clinicians are expected to encounter small TP53 ^mut clones of uncertain significance. Studies reporting outcomes used variable thresholds of TP53 ^mut VAF (ranging from 1% to 20%) [2, 6, 8, 10]. Therefore, additional clarity is needed to determine if an optimal TP53 ^mut VAF reliably distinguishes CH/CCUS from MN and those with poor survival.

Finally, at the peril of stating the obvious, ultimate progress will be measured by offering safe, effective, and durable treatments. The collective approach would include early institution of therapy and achieving higher response rates—potentially increasing the patients eligible for alloHCT.

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

M.V.S. receives research funding to the institution from Astellas Pharma, AbbVie, KURA Oncology, Inc., BMS/Celgene, and Marker Therapeutics. D.K.H. is a member of the board of directors or advisory committees of AbbVie and Novartis.

Data Availability Statement

The authors have nothing to report.