A line in shifting sand: Can we define and target TP53 mutated MDS?

Abstract

Mutations in the tumor suppressor protein, TP53, lead to dismal outcomes in myeloid malignancies, including myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). Recent pathological reclassifications have integrated TP53 mutated MDS and AML under a unified category of TP53 mutated myeloid neoplasms, which allows for more flexibility in treatment approaches. Therapeutic strategies have predominantly mirrored those for AML, with allogeneic stem cell transplantation emerging as critical for long-term disease control. The question remains whether there are physiological distinctions within TP53 mutated myeloid neoplasms that will significantly impact prognosis and therapeutic considerations. This review explores the unique aspects of classically defined “TP53 mutated MDS”, focusing on its distinct biological characteristics and outcomes. Our current understanding is that TP53 mutated MDS and AML are globally quite similar, but as a group have unique features compared to TP53 wildtype (WT) disease. Optimizing immunotherapy and targeting vulnerabilities due to co-mutations and/or chromosome abnormalities should be the focus of future research.

Introduction

Across all cancer types, TP53 mutations are associated with a poor prognosis, especially for TP53 mutated myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML). Historically, the distinction between the 2 entities has relied on blast percentage with 20% blasts being the cutoff from MDS to AML. Recently, the pathologic classification in the World Health Organization (WHO) and newly formed International Consensus Classification (ICC) have added a separate category for TP53, blurring the definition. ICC characterizes as ‘myeloid neoplasms with TP53’ and WHO has added ‘MDS with biallelic TP53 inactivation’ [1,2]. With subtle nuances to each of the categories, understanding biology, pathophysiology, and treatment of this disease has become more challenging. Furthermore, clear definitions are essential for optimal design and implementation of clinical trials to advance care. A looming unanswered question is whether TP53 mutated MDS and AML should be considered separate processes or if TP53 mutated myeloid malignances should be a homogenous category and therefore blast agnostic. While there are many similarities between TP53 mutated MDS and AML as recently described in several detailed reviews [3–6], the focus of this review will be on understanding what makes TP53 mutated MDS unique.

TP53 Mutated MDS Outcomes

Epidemiology

Alterations in TP53 are identified in 5%–10% of de novo MDS or AML cases [7–13], with this number increasing to 20%–40% in patients with therapy-related myeloid neoplasms [14–16]. Alterations include TP53 mutations, chromosomal abnormalities leading to loss of a TP53 gene or copy-neutral loss of heterozygosity (cn-LOH) at the TP53 locus [7,17–20]. The majority of TP53 aberrations are due to missense mutations in the DNA binding domain (75%) [21]. Additional TP53 aberrations are deletions and frameshift insertions (9%), nonsense mutations (7%), silent mutations (5%) and other rare alterations (2%) [11,12,22,23]. 66%–76% of patients have “multihit” or “bi-allelic” TP53 mutational status, defined as 2 or more distinct TP53 mutations with variant allele frequency (VAF) greater than or equal to 10%, a single TP53 mutation associated with either TP53 cytogenetic deletion or cn-LOH, or a single TP53 mutation detected with a VAF greater than 55% [12,13]. The VAF threshold of 10% was established to mitigate the risk of mistakenly attributing a poor prognosis to a nonleukemic clonal hematopoietic variant.

Prognosis

TP53 mutated myeloid neoplasms represent a significant unmet need in the field with dismal median overall survival (OS) rates of 5 to 10 months [13]. Irrespective of disease state being MDS or AML, OS is worse for TP53 multihit allelic status compared to mono-allelic state [24] with an increased risk of leukemic transformation and a 5-year OS of less than 6% [12]. Multihit state also correlates with increased likelihood of complex karyotype (CK), which is defined as presence of 3 or more chromosomal abnormalities and portends a negative prognosis independent of co-mutations. When CK is identified, the presence of multiple somatic mutations is less commonly seen [11]. This is dependent on the allelic state of TP53 mutations, as in multi-hit TP53 or biallelic cases; in these situations TP53 mutations are early events, evolve to become the dominant clones, and less frequently have concurrent mutations [12]. Conversely, 90% of monoallelic cases harbor co-mutations. Importantly, one study of monoallelic TP53 mutation without other mutations determined that this subtype had a 5-year OS that is similar to TP53 WT MDS at 81% compared to 36% with 1 or 2 co-mutations, and 26% with 3–4 mutations [12]. Importantly, over 80% of these monoallelic TP53 cases had a non-complex karyotype and may be enriched for isolated monosomy 5 or deletion 5q. As discussed below, TP53 mutated MDS with isolated monosomy 5 or deletion 5q may represent a biologically distinct disease with improved survival [19,25]. The consensus is that patients with TP53 mutated MDS patients that are multihit and/or complex karyotype have dismal outcomes.

A numerically higher TP53 variant allele frequency (VAF) has also been associated with an unfavorable prognosis. VAF >40% are more likely to have CK and less frequently have additional mutations [26]. In TP53 mutated MDS specifically, presence of high VAF defined as greater than 40% is associated with worse OS but not event-free survival (EFS). This finding is not seen when VAF is less than 20% nor when VAF is between 20% and 40% [27]. Despite some studies showing an association between VAF and OS, other studies do not show any impact on VAF thresholds and OS or EFS [13,26].

Until 2022, the Revised International Prognostic Scoring System (IPSS-R) was the primary risk stratification tool used at MDS diagnosis to predict prognosis. The updated Molecular International Prognostic Scoring System for MDS (IPSS-M) incorporates molecular mutations in addition to conventional cytogenetics [28]. The number of TP53 mutations with maximum VAF and presence or absence of loss of heterozygosity are now included. The newer model more accurately predicts risk of transformation and is useful for treatment decisions including whether patients should consider allogenic hematopoietic stem cell transplantation (HSCT).

Testing

Testing for TP53 mutations can be done by cytogenetics, fluorescence in situ hybridization (FISH) testing and next generation sequencing (NGS). All 3 have advantages and disadvantages with regards to turnaround time, sensitivity and type of mutation detected. Clinically, irrespective of the type of mutation, presence or absence remains the most vital for tailoring treatment decisions. Detection of TP53 mutations by immunohistochemistry (IHC) staining for p53 protein expression is of growing interest and importance for risk stratification, management of treatment and decision for HSCT. The turnaround time of 1–3 days is particularly appealing and, in most cases, can be done locally after internal pathology validation. Overall, p53 IHC appears to correlate with TP53 mutational status, however, there are different staining patterns based on type of mutation [29–31]. Missense mutations, which are the most common mutation seen, have a ‘positive’ stain defined as a greater than or equal to 10% 3+ staining. Large deletions, nonsense mutations, or intronic splice-site variants that result from truncated protein will stain as ‘negative or ‘null’ patient and would also be considered ‘positive’ for a TP53 mutation [32–34]. Details of testing patterns are outside the scope of this review. Three groups have correlated p53 IHC staining with TP53 mutations with similar sensitivity and specificity in AML [35–37]. Additional studies are needed to further validate p53 IHC staining in myeloid malignancies with lower blast percentage.

Treatment

The majority of treatment options and studies published to date have focused on TP53 mutated AML. It is no surprise that this disease is challenging to treat; the distinction between MDS and AML treatment for these patients provides an additional layer of complexity. Based on evolving information, some of which is presented here, we propose that treatment of TP53 MDS and AML be treated as one entity under the umbrella of TP53 myeloid malignancies. While the underlying biology and pathophysiology of MDS and AML is distinct for certain subgroups, it is becoming increasingly clear that the presence of TP53 mutations supersedes blast percentage. Regardless, data and therefore discussion of current recommendations specific for TP53 mutated MDS is limited.

Intensive chemotherapy is not a current treatment strategy for classically defined TP53 mutated MDS as it has yielded disappointing results with low complete response (CR) rates, OS, and increased toxicity. The biggest decision at diagnosis should involve the discussion of whether the patient is an allogenic HSCT candidate. Subsequently, presence or absence of blasts will then determine the ideal short-term treatment plan to optimize the patient for HSCT. For patients with TP53 mutated MDS with no excess blasts, chemotherapy prior to HSCT may not be needed. Chemotherapy introduces toxicities and may increase the time it takes to get to transplant. When blast percentage is above 10%, it is generally accepted that the blasts need to be reduced to less than 5% to proceed to HSCT. However, the data supporting this approach to HSCT timing is limited to retrospective and single arm studies [38–42]. The VidazaAllo study assessed the benefit of 4–6 cycles of hypomethylating agent (HMA) prior to HSCT in elderly, high-risk MDS patients and found 33% of patients could not proceed to HSCT due to disease progression, drug-related adverse events or new comorbidities [38,43]. Thus, the key is to get to HSCT as quickly and safely as possible.

For older patients, it is reasonable to start with single agent HMA such as decitabine or azacitidine [44,45]. As a single agent, it may take 4–6 cycles to see a response, but overall treatment tends to be well tolerated. For younger patients, consideration of adding venetoclax is worth discussing but is not yet FDA approved [46]. The Phase III VERONA trial for newly diagnosed MDS combines azacitidine with 14 days of venetoclax (as opposed to 28 days in AML) with improved response rates. Blast clearance tends to be faster, occurring within 1–2 cycles. However, the cytopenias are much more significant, which increases bleeding risk and potential for infectious complications. In TP53 mutated AML patients with complex cytogenetics treated with azacitidine and venetoclax on VIALE-A, response rate was higher compared to azacitidine alone, but duration of response was similar at 6 months for both groups [47]. Conversely, Badar et al. [48] has showed increase duration of response with azacitidine and venetoclax compared to azacitidine alone in TP53 mutated AML, but EFS and OS was similar between the 2 treatment groups. Extrapolating to the TP53 mutated MDS, it is not clear that venetoclax is beneficial in patients not going to HSCT. Unfortunately, limited options and lack of response to novel therapies (discussed below) highlight the importance of persisting in attempt to improve outcomes for this patient population. As noted, we feel that if patients are eligible for allogeneic stem cell transplant, initial treatment decisions should be based on optimizing the bridge to transplant.

Biological Differences in TP53 Mutated MDS versus AML

As described above, TP53 mutant MDS and AML have similarly poor response to current therapies. The question remains whether there are biologic differences between TP53 mutated MDS and AML that would indicate different approaches to treatment. Intact p53 functions in 3 broad categories, including (1) regulation of DNA damage/repair, cell cycle arrest, apoptosis, senescence and autophagy, (2) regulation of metabolic pathways and broad control over catabolic versus anabolic processes, and (3) regulation of the tumor microenvironment [49]. Literature suggests biological differences of MDS versus AML on the above pathways [50]. Thus, we will first discuss the data comparing the physiological dependences of TP53 mutated MDS versus AML on these p53 functions.

While several studies from our group and others have evaluated p53 functions in TP53 mutated AML, there is a paucity of data that has addressed these biological consequences in MDS and even less data comparing the 2 “diseases” head-to-head. One of the only studies that has directly compared TP53 mutated MDS, AML, AML with myelodysplasia-related changes (AML-MRC) and therapy-related AML identified differences in mutation types, spectrum and hot spots between disease categories [51]. Consistent with prior reports, missense mutations in the DNA-binding domain were most common across all categories. Interestingly, inactivating mutations and mutations outside the DNA binding domain were more common in AML-MRC than MDS. One of the most striking findings was that TP53 mutations in MDS were more likely to retain transcriptional activity [51], suggesting a potential biological difference though not one that has significantly altered response to standard therapies to date.

With regard to canonical functions of p53, particularly in apoptosis, there is significant clinical data to suggest that TP53 mutated MDS and AML have equally poor responses and resistance to Bcl2 inhibition by venetoclax [47,48]. Preclinical data evaluating the mechanism of venetoclax resistance in TP53 MDS is less studied than in TP53 mutated AML, where data suggests alternative BH3 family proteins and mitochondrial metabolism as key mechanisms of drug persistent cells [52–54]. Additional preclinical studies from our group and others support the role of mitochondrial metabolism and its intersection with the cholesterol synthesis pathway as essential for TP53 mutated AML, but have not been evaluated in TP53 mutated MDS models [55–58].

Most of the work that directly compared TP53 mutated MDS and AML has been regarding the role of p53 in immune regulation. Multiple studies have demonstrated that TP53 mutated MDS and AML patient samples have evidence of an immune-privileged, evasive phenotype that may be driving poor outcomes [59–61]. Gene expression analyses demonstrated that TP53 mutated, compared to wildtype, AML bone marrow samples, are enriched for interferon gamma, immune checkpoints, and immune senescence [59]. Work from Sallman et al. [60] demonstrated that bone marrows from both TP53 mutated MDS and AML patients have reduced cytotoxic and helper T cells with a concurrent increase in highly immunosuppressive regulatory T cells and myeloid-derived suppressor cells compared to TP53 wildtype disease. Furthermore, a higher proportion of regulatory T cells was a significant independent predictor of poor overall survivor in this cohort [60]. Zeidan et al. [61] compared separate MDS and AML cohorts of TP53 mutated versus wild-type disease with a particular focus on the immune landscape, as characterized by flow cytometry and bulk RNA sequencing techniques. Similar to prior studies, TP53 mutated patients had a higher T cell population and upregulation of inhibitory immune checkpoint proteins, like programmed death-ligand 1 (PDL1), on CD34+ AML blasts. Furthermore, AML patients with TP53 mutations had a higher abundance of exhausted T cells compared to TP53 wild-type AML patients [61]. An independent study suggested TP53 mutated AML cells engaged by CAR T-cells have a prolonged interaction time associated with upregulation of exhaustion markers, with TP53 mutated cells exhibiting upregulated mevalonate pathway and downregulated Wnt pathway activity compared to TP53 wildtype cells [57]. Other studies suggest exhaustion induced by direct cytokine release from AML cells themselves [62]. Regardless, while TP53 mutated AML and MDS were not directly compared, it is notable that the differences were most striking between TP53 mutated versus wildtype AML whereas there were no statistically significant differences in percentage of T cells, myeloid progenitors, or PD-L1 expression between TP53 mutated vs wildtype MDS [61]. This data indirectly suggests that TP53 mutations in myeloid malignancies alter the immune environment and that the TP53 mutated MDS and AML immune environment is more similar to each other and other TP53 wildtype MDS patients than to TP53 wildtype AML patients. This data raises an interesting but still un-addressed question as to whether loss of immune suppression regulates the progression from TP53 mutated MDS to AML, rather than a cell intrinsic change.

Importantly, we may gain more insight from identifying key biological differences based on the concurrent mutations and karyotype abnormalities in TP53 mutated MDS/AML patients. Demonstration of distinct co-mutation/karyotype abnormalities between TP53 mutated myeloid subtypes has been described [51,63]. Co-mutations in 26 driver genes are underrepresented in TP53 non-sense or splice mutations compared to TP53 missense mutations [51]. Myelodysplasia-related gene co-mutations (25% of the 130-patient cohort), such as ASXL1 (16%), BCOR, EZH2, RUNX1, SF3B1, SRSF2, STAG2, U2AF1, and/or ZRSR2 are associated with a more favorable survival in TP53 mutated AML (HR 0.366, 95% CI, 0.181–0.738, P = .005) [64]. Conversely, TP53 bi-allelic loss in JAK2-V617F myeloproliferative neoplasms is associated with exposure to chronic inflammation, transformation to erythroid-biased secondary AML, and dismal outcomes [65–67]. Unfortunately, in neither case, do we have insight into how cooperating mutations impact the biology of the resulting disease.

Co-occurrence of TP53 mutations with aberrations of chromosome 5 is one of the most studied entities within the TP53 mutated myeloid neoplasm spectrum. A study of 318 MDS patients identified chromosome 5 aberrations in all patients with TP53 mutations [19]. Notably, 20% of patients had isolated deletion 5q (del5q) while the remainder had complex karyotype with monosomy 5 or del5q [19]. Patients with isolated del5q MDS have similar outcomes regardless of TP53 mutation status and longer overall survival compared to TP53 mutated MDS with monosomy 5/5q-in the context of complex karyotype [25], suggesting these 2 TP53 MDS subtypes are biologically distinct and could be treated differently.

Insight into the biological role of TP53 mutations in isolated del5q MDS was identified in a model in which hematopoietic stem and progenitor cells (HSPCs) with del5q were negatively impacted by inflammation, which has been described as a contributor to MDS in other systems [68]. Del5q HSPCs exposed to inflammation became less quiescent, which raises the concern for subsequent exhaustion. The decrease in quiescence was reversed by loss of p53, suggesting a competitive advantage for TP53 mutated, del5q MDS cells [68]. The advantage of p53 loss in del5q MDS has been validated in mouse models with loss of key chromosome 5q genes, such as nucleophosmin 1 (NPM1) and casein kinase I isoform alpha (CSNK1A1) [69,70]. Notably, del5q clones with concurrent loss of p53 may also undergo selective pressure from conventional therapies such as lenalidomide, used in the treatment of isolated del5q MDS. Lenalidomide leads to p53-mediated apoptosis of del5q cells due to haploinsufficient expression of CK1a, which is the protein encoded by CSNK1A1. Thus, loss of p53 can alter lenalidomide responses and may increase the risk of leukemic transformation [17,18,71,72].

While our current understanding of co-mutations/aberrations is largely limited to their impact on prognosis, the goal is to determine how these molecular abnormalities can lead to targetable vulnerabilities despite the concurrent TP53 mutations. A hopeful development is the increased sensitivity of monosomy 7/del7q MDS/AML to nicotinamide phosphoribosyltransferase (NAMPT) inhibitors due to NAMPT haploinsufficiency that is independent of TP53 status [73,74]. Current and future therapeutic strategies to target these vulnerabilities will be discussed below.

Current and Future Therapeutic Strategies in TP53 Mutated MDS

As common as the mutation of TP53 is in cancer, the failure of TP53 targeting agents is even more common. There have been 2 major therapeutic approaches to date. First has been reactivation of p53. The second is immunotherapy. Both approaches have tried to restore apoptosis and both approaches have phase III failures (Table 1).

Table 1.

Summary of prior seminal trials in TP53 mutated MDS.

Treatment Approach	Mechanism of action	Control Arm	Population	Trial Size	Design	TP53 Response	TP53 median OS (months)	Trial Identifier	Status
Venetoclax (400 mg PO daily, days 1–14) + azacitidine (75 mg/m2 SQ D1–7 q28 days)	BCL2 inhibitor + HMA	azacitidine (75 mg/m2 sq D1–7 q28 days)	Int to HR-MDS	planned n = 530; no pre-determined TP53 stratification	Phase III - Randomized open label, double arm Phase Ib-Open label single arm	ORR: N/A CR: N/A (80% and 30% in entire Phase Ib cohort)	N/A (26 months in entire Phase Ib cohort)	NCT04401748 VERONA	Ongoing
Pevonedistat (20 mg/m2 IV D1,3,5) + azacitidine (75 mg/m2 IV or SQ on days 1 to 5, 8, and 9)	NEDD8 inhibition + HMA	azacitidine (75 mg/m2 on days 1 to 5, 8, and 9)	HR-MDS, CMML, AML	n = 454 AML / MDS / CMML; 39 (24%) TP53 MDS in experimental arm	Phase III - multicenter, open-label, double arm	ORR: 25% CR: N/A	TP53: 21.6	NCT03268954 PANTHER	Completed, development stopped
Magrolimab (IV 1 mg/kg then ramp to 30 mg/kg once-weekly) + azacitidine (75 mg/m2 IV or SQ daily days 1–7)	CD47 inhibtion + HMA	placebo + HMA	Int to HR-MDS	planned n = 530; prior phase Ib n = 95; 25 (26%) TP53 in experimental arm	Phase III - Randomized, open-label, double arm Prior phase Ib-open label, single-arm	PhIB available. ORR: 68% CR: 40%	TP53: 16.3 Entire cohort: Not reached	NCT04313881 - ENHANCE	Stopped Early
Durvalumab (1500mg IV q28 days) + azacitidine (75 mg/m2 SQ D1–7 q28 days)	PD-L1 Inhibition + HMA	azacitidine (75 mg/m2 sq D1–7 q28 days)	Int to HR-MDS	n = 84; 14 (33%) TP53 in experimental arm	Phase II - Randomized, open label, double arm	ORR: 41% CR: N/A	TP53: N/A Entire cohort: 12	NCT02775903	Completed
Eprenetapopt (4.5 g IV days 1–4 every 28 days cycle) + azacitidine (75 mg/m2 SQ D1–7 q28 days)	p53 stabilizer + HMA	azacitidine (75 mg/m2 sq D1–7 q28 days)	TP53mt high risk MDS	n = 154; 78 (100%) TP53 in experimental arm	Phase III - Randomized open label, double arm	ORR: N/A CR: 33%	N/A (10.8 months in prior Ph II)	NCT03745716	Stopped Early
Eprenetapopt (3.7 g IV daily infusion days 1–4 every 28 days cycle) + Azacitidine (37.5 mg/m2 days 1–5)	p53 stabilizer + HMA	azacitidine (37.5 mg/m2 sq D1–5 q28 days)	TP53 MT MDS/AML (maintenance post-SCT)	n = 55 AML/MDS; 19 TP53 MDS in experimental arm	Phase II - Open-label, single arm trial	N/A	TP53: 20.6	NCT03931291	Completed
Sabatolimab (400mg IV on day 8 and 22) + HMA (decitabine 20 mg/m2 IV on day 1–5 or azacitidine 75 mg/m2 IV or SQ on day 1–7 or day 1–5 and day 8 and 9)	TIM Inhibtion + HMA	placebo + HMA	Int to HR-MDS	n = 127, 22 (35%) TP53 in experimental arm	Phase II - Randomized, double-blind, double arm	ORR: 71% CR: N/A	TP53: N/A Entire cohort: 19	NCT03946670 & NCT04266301	Stopped Early

Abbreviations: AML = acute myeloid leukemia; Bcl2 = B-cell leukemia/lymphoma 2; CMML = chronic myelomonocytic leukemia; CR = complete response; D = day; HMA = hypomethylating agent; HR = high risk; int = intermediate; IV = intravenous; MDS = myelodysplastic syndrome; n = sample size; NEDD8 = neural precursor cell expressed developmentally downregulated protein 8; ORR = overall response rate; OS = overall survival; PD-L1 = programmed cell death ligand 1; PO = per oral; q = cycle frequency; SCT = stem cell transplant; SQ = subcutaneous; TIM = T-cell immunoglobulin and mucin domain.

Loss of p53 function has a profound survival advantage for clonal cells. Consequently, reactivation of p53 has drawn strong interest in clinical development. Eprenetapopt (APR-246) had been the lead drug candidate among the class of small molecules intended to restore wild-type p53 functions in TP53-mutated cells. In a Phase I/II trial (NCT03072043) of eprenetapopt and azacitidine, 40 patients with at least 1 TP53 mutation received combination therapy. 50% achieved a complete response as well as reduction in both TP53 variant allele frequency and IHC p53 staining. Median overall survival was 10.8 months [75]. Unfortunately, a randomized phase III trial (NCT03745716) failed to demonstrate a statistically significantly higher complete response for combination therapy versus azacitidine monotherapy (33% versus 22%, P = .13). Nonetheless, there remain promising efficacy signals for eprenetapopt. Uncontrolled trials using eprenetapopt and reduced dose azacitidine as maintenance therapy following allogeneic HSCT and triplet combination therapy of venetoclax, azacitidine and eprenetapopt in TP53 mutated AML both showed promising response rates and median overall survival compared to historical expectations. Randomized trials would need to confirm clinical benefit [76,77]. Additionally, novel combinations to overcome resistance to eprenetapopt such as eprenetapopt and nuclear export proteins or combination p53 reactivating small agents may offer additional opportunities for development [78].

Given the challenges of restoring p53 function, the other major strategy has been harnessing the immune system to target TP53 mutated cell resistance to classical cytotoxic agents (Fig. 1). Until recently, the most promising approach had been restoring innate immune function via CD47 blockade with the monoclonal antibody. CD47 is the ligand for signal-regulatory protein (SIRP)α and activation of those pathways leads to antiphagocytotic signals. Preclinical studies have shown that CD47 blockade is not sufficient for phagocytosis of tumor cells. For antitumor activity, another co-stimulatory signal is needed. Tumor cells may need to express additional prophagocytic markers. These may include signaling lymphocytic activation molecule family (SLAMF)7 with macrophage-1 antigen (Mac-1), calreticulin induced by azacitidine, additional activation of macrophages via Fcγ receptor activation by the Fc region of a monoclonal antibody targeting CD47 or by a second agent targeting leukemia cells (eg, CD33). The first agent targeting CD47 entered clinical trials in 2014. There are now over 20 agents in active clinical development across 30 trials spanning multiple indications and cancer types [79]. The leading approach in TP53 mutated MDS had been magrolimab combined with azacitidine. In a phase I study (NCT03248479) of 25 TP53 mutated MDS patients, 40% achieved CR with median OS of 16.3 months. Thirty-four patients (36%) had allogeneic HSCT with 77% 2-year OS [80]. Unfortunately, the ENHANCE-2 study, a double-blind placebo controlled randomized phase 3 study of azacitidine and magrolimab versus azacitidine and venetoclax was stopped for futility. Similarly, therapies like durvalumab and sabatolimab, directed at the T-cell immunoglobulin and mucin domain (TIM)-3, PD-L1 and programmed cell death (PD)1 axis of the adaptive immune system, have also failed [81] (NCT04266301). In considering these studies, it is important to note that full characterization of the immune response to AML cells is lacking and these agents may still have a role to play in combination with other immune targeting approaches.

Fig. 1. Ongoing drug development targets in high risk MDS.

Schema of active targets for drug development in high risk MDS, including TP53 mutated myeloid neoplasms with a focus on cellular immunotherapy and drugs targeting cell signaling, epigenetics, gene expression, cell cycle, apoptosis, p53 activity, proteosome activity and metabolic pathways, including mitochondria, cholesterol and NAD synthesis. Abbreviations: TGF = transforming growth factor; FLT3 = fms-like tyrosine kinase 3; ATR = ataxia telangiectasia and Rad3-related protein; PRTM5 = protein arginine methyltransferase 5; DNA = deoxyribonucleic acid; WNT = wingless/integrated; HMG-CoA = hydroxymethylglutarylcoenzyme A; NF-kB = nuclear factor kappa B; IKK = inhibitor of nuclear factor kappa B; DMAPT = dimethylamino parthenolide; NEDD8 = neural precusor cell expressed developmentally down-regulated 8; BiTE = bispecific T-cell engager; ADC = antibody drug-conjugate; SIRPa = signal regulatory protein alpha; TIM3 = T-cell immunoglobulin and mucin domain 3; IDH = isocitrate dehydrogenase; TP53 = tumor protein 53; MDM2 = murine double minute 2; NAMPT = nicotinamide phosphoribosyltransferase; BCL = B-cell leukemia/lymphoma.

The remaining pipeline agents focus on immune therapies that do not rely on a functional apoptosis system (eg, pivekimab and flotetuzumab), targeting retinoic acid receptor alpha (RARA) overexpression (eg, tamibarotene), or even novel conditioning therapies as part of allogeneic HSCT (eg, iomab-b). Novel approaches increasingly focus on the need to overcome metabolic drivers of therapy resistance in TP53 mutated MDS or AML cells, such as the mitochondrial stress response and cholesterol synthesis [55–57,82]. TP53 mutated MDS or AML cells often have multiple chromosomal abnormalities, including monosomy 7, which may confer vulnerability to NAMPT inhibition and the subsequent decrease in energy carrier, NAD [73,74]. Furthermore, wingless-related integration site (WNT) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB) are essential and targetable in TP53 mutated myeloid neoplasms [57,83,84]. Monoclonal antibodies, T-cell engager antibodies, alloreactive natural killer (NK) and chimeric antigen receptor (CAR)T-cell therapy, immune checkpoint blockade beyond PD-1/PD-L1 or cytotoxic T-lymphocyte associated protein (CTLA)-4, such as TIM3, are still actively being investigated. Even when a target exists, all immunotherapy approaches seem to face hurdles of collateral hematopoietic damage or inactivity. If such approaches can work in MDS/AML, they may very well work in TP53 mutated myeloid disease as well.

Conclusion

In conclusion, TP53 mutations represent a significant challenge in MDS and AML due to their profound impact on treatment response and overall outcomes. Current studies indicate that TP53 mutated MDS and AML are biologically similar but may exist on a disease spectrum with progressive loss of p53 function from MDS to AML given the differences in mutation types, spectrum and VAFs between disease categories. However, the line in the sand appears to be appropriately drawn between TP53 wildtype and mutated myeloid neoplasms, as even the loss of p53 function appears to have a significant, negative impact on outcomes. We support the integration of TP53 mutated MDS and AML into a unified category of TP53 mutated myeloid neoplasms particularly as it may facilitate a more flexible approach to treatment. However, there may still be unique features across the disease spectrum that could identify distinct therapeutic opportunities. Future research should focus on delineating these differences more clearly, optimizing immunotherapy, and identifying vulnerabilities related to co-mutations and chromosomal abnormalities. By advancing our understanding of these nuanced aspects, we can improve risk stratification, enhance therapeutic strategies, and ultimately, provide more tailored and effective treatments for patients with TP53 mutated myeloid malignancies.