The Prognostic Value of TP53 Mutations in Adult Acute Myeloid Leukemia: A Meta-Analysis

Abstract

Objective

Mutations of the tumor protein p53 (TP53) gene were considered to be associated with an unfavorable prognosis in acute myeloid leukemia (AML). This meta-analysis aimed to systematically elucidate the prognostic value of TP53 mutation in adult patients with AML.

Method

A comprehensive literature search was conducted for eligible studies published before August 2021. The primary endpoint was overall survival (OS). Pooled hazard ratios (HRs) and their 95% confidence intervals (CIs) were calculated for prognostic parameters. Subgroup analyses based on intensive treatment were performed.

Results

Thirty-two studies with 7,062 patients were included. As compared to wild-type carriers, AML patients with TP53 mutations had significantly shorter OS (HR: 2.40, 95% CI: 2.16–2.67, I²: 46.6%). Similar results were found in DFS (HR: 2.87, 95% CI: 1.88–4.38), EFS (HR: 2.56, 95% CI: 1.97–3.31), and RFS (HR: 2.40, 95% CI: 1.79–3.22). Mutant TP53 predicted inferior OS (HR: 2.77, 95% CI: 2.41–3.18) in the intensively treated AML subgroup, compared with the non-intensively treated group (HR: 1.89, 95% CI: 1.58–2.26). Among intensively-treated AML patients, the age of 65 did not affect the prognostic value of TP53 mutations. Besides, TP53 mutation was also strongly associated with an elevated risk of adverse cytogenetics, which conferred a dismal OS in AML patients (HR: 2.03, 95% CI: 1.74–2.37).

Conclusion

TP53 mutation exhibits a promising potential for discriminating AML patients with a worse prognosis, thus being capable of serving as a novel tool for prognostication and therapeutic decision-making in the management of AML.

Introduction

Acute myeloid leukemia (AML) is a genetically heterogeneous disorder characterized by the somatic acquisition of genetic and epigenetic alterations in hematopoietic progenitor cells that perturb normal self-renewal, proliferation, and differentiation mechanisms [1]. Of acute leukemia in adults, AML is the most common, with an incidence exceeding 20,000 cases per year in the USA alone [2]. The advent of sequencing techniques such as next-generation sequencing has allowed new insights into the molecular basis of AML, and the molecular analysis of gene alterations has been incorporated into the risk stratification and prognostication for patients with AML in clinical practice [3].

The tumor suppressor p53 (TP53) gene, located on chromosome 17p13.1, is an established cellular gatekeeper of proliferation/differentiation in response to aberrant oncogene expression. TP53 inactivation induced by gene-mutation or deletion favors the effect of the oncogene, thus contributing to the uncontrolled proliferation of cancer cells [4]. Somatic mutations of TP53 gene have been detected in more than 50% of all human tumors and represent an important prognostic and predictive marker in cancer [5]. Of note, these mutations are less frequent in hematological malignancies (10% overall); however, similar to other cancer types, they are found to show a negative impact on survival [6]. In patients with myelodysplastic syndromes, TP53 mutations (5–10% of cases) are associated with high-risk disease, secondary AML progression, and dismal outcomes [7]. Similarly, TP53 mutations are observed in proximately 5–10% of adult de novo AML cases [8], with high frequency (∼30%) in therapy-related AML. TP53 mutations are also highly associated with increased ages and complex karyotypes (CK) (52–70% of cases) [9, 10, 11].

Quite a few reports have supported the adverse prognostic impact of TP53 mutations in the setting of AML, where patients with mutant TP53 could exhibit a significantly shorter overall survival (OS) than those with wild-type TP53. The 2017 European Leukemia Net guidelines were updated to recommend upfront screening for mutations in TP53 at diagnosis for risk stratification given its dismal prognostic implications [3]. Given the relatively low frequency of TP53 mutation in this population, additional information is encouraged for establishing a prognostic profile of TP53 mutations based on previously published data in a systematic manner. Hence, this present meta-analysis was conducted to systematically validate the prognostic value of TP53 mutations in adult patients with AML through comparing the prognostic parameters between patients with mutant and wild-type TP53 to strengthen the impact of this gene mutation in clinics, and the relations between TP53 mutation status and cytogenetics would also be assessed.

Materials and Methods

Data Sources and Search Strategy

The present study was conducted following the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statements [12]. A database search of PubMed, Cochrane library, ScienceDirect, Embase, and Web of Science was conducted to find eligible studies published before August 2021. The following search terms with their synonyms in various combinations were used in the searching process: “Myeloblastic,” “Leukemia,” “TP53,” and “prognosis.” Reference lists of potentially related papers were cross-checked to identify additional records. After removing duplicate items with the help of Endnote X9 software, two reviewers independently screened the initial search results based on title and abstract and then, selected candidate studies for further full-text evaluation. Any difference was resolved by consensus.

Eligibility Criteria

Studies were considered eligible if satisfied the following requirements: (1) prospective or retrospective clinical studies assessing TP53 mutation status in adult patients with AML; (2) offered enough prognostic information of TP53 mutation versus TP53 wild-type subgroups, with calculation of hazard ratio (HR) and its corresponding 95% confidence intervals (CI) or with adequate data for their estimation; and (3) original articles written in English. Those articles were excluded if they had a mixed cohort with <50% of the cohort having AML without a separate analysis of AML cases. In event of overlapped samples, the study with the most sufficient data was included. Nonrelated papers, animal studies, reviews, case reports, letters, editorials, conference abstracts, duplicate publications, or studies with insufficient data were rejected from the meta-analysis.

Data Extraction

For each eligible study, we captured the following information: the first author, year of publication, country, study design, sample size, patient age, type of AML, median follow-up, therapeutic strategies, TP53 mutation status (measuring techniques and incidence of mutations), and outcomes of interest including HR for prognostic information with its 95% CI under Cox proportional hazards regression models, such as OS (the time interval from the date of diagnosis to the date of death), RFS (the time interval from the date of complete response to the date of relapse), event-free survival (EFS, defined as the time from diagnosis of AML to treatment failure relapse, death, or last follow-up), and disease-free survival (DFS, the duration from complete response until relapse or death or censorship at the last follow-up). All data were checked to ensure the study's accuracy and any discrepancies were finally resolved by consensus.

Quality Assessment

Two reviewers independently assessed the quality of each eligible study with NOS (Newcastle-Ottawa Quality Assessment Scale) [13] and reached a consensus by discussion in cases of differences. NOS covers three domains: selection (four items), comparability (two items), and outcome (three items). The overall NOS score ranges from zero to nine points. Studies with more than six points were considered to be of high quality.

Statistical Analysis

Pooled HRs and their 95% CIs for OS (the primary objective) comparing TP53 mutation and TP53 wild type were used to assess the prognostic value of TP53 mutations in patients with AML. Other prognostic parameters, such as RFS, EFS, and DFS, were secondary outcomes. The HR >1 with 95% CI exceeding 1 represents a significantly higher risk of death in favor of AML patients with TP53 mutations, compared to those with wild-type TP53. Additionally, odds ratio (OR) and 95% CI were used to reveal relations between TP53 and patients with adverse cytogenetics, with an OR and 95% CI >1, suggesting that patients with mutant TP53 were at a higher risk of adverse cytogenetics with statistical significance. Heterogeneity across studies was measured by the inconsistency index (I²), and an I² >50% was indicative of significant heterogeneity [14]. We used a random-effects model for data pooling when I² >50% was present; otherwise, the fixed-effect model was used [14]. Sensitivity analysis was conducted to interpret significant heterogeneity by sequentially omitting 1 study each time. Besides, subgroup analyses based on the treatment strategies (intensively treated vs. nonintensively treated) were performed for the primary outcome (OS) to investigate the impact of intensive treatment on the overall result. Studies in which AML patients treated with intensive treatment accounted for more than half (>50%) of the total cohort were grouped into the “intensively treated” subgroup. Further analysis based on median age (≥65 years old or not) within each subgroup was done to investigate the impact of age. A test of interaction was used to explore if there was a statistical difference (p < 0.05) between subgroups [15]. The potential publication bias of the primary outcome was evaluated through Egger's test, with a p > 0.05, indicating the absence of potential bias [14]. A trim-and-filled analysis was performed if there was a p < 0.05. All the data analyses were conducted by STATA version 12.0 (College Station, TX, USA).

Results

Literature Search

In total, 7,314 relevant studies were identified from electronic databases and cross-checked references. After excluding duplicated records, we screened the results based on title and abstract and selected 179 studies for further full-text evaluation. Finally, 32 studies [16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47] were included in our meta-analysis. Figure 1 shows the flowchart of study selection.

Fig. 1.

Flow diagram of study selection.

Study Characteristics and Quality Assessment

Published between 2002 and 2021, 31 studies including 7,062 subjects from eight countries were included in our study. Fourteen articles (43.8%) [16, 20, 21, 22, 23, 24, 25, 29, 32, 35, 36, 37, 38, 42] were prospective studies and the remaining were in a retrospective design. The median age of enrolled AML cases in 14 studies was ≥65 years old. More detailed information is presented in Table 1. The NOS was used for quality assessment and results are summarized in Table S1 (see www.karger.com/doi/10.1159/000526174 for all online suppl. material). The average NOS score across these studies was 6.1 (Table 1), suggesting that the methodological quality was moderately high and overall acceptable.

Table 1.

Study characteristic of the included studies

First author	Year	N	Median age	M, %	Type of AML	Median OS/FU	NOS
Bories	2020	279	76	55.6	De novo AML, sAML, tAML	FU 66.1 months	7
Chen	2021	204	54.4	50.5	De novo AML	/	5
Daher-Reyes	2021	435	59	52.0	De novo AML, sAML, tAML	FU 475 days	8
Falk	2015	189	64	50.3	De novo AML	/	4
Fenwarth	2019	187	(18–59)	52.0	De novo AML	/	7
Fernandez	2017	143	60	50.3	AML-MRC, AML-NOS, AML-RGA, tAML	/	5
Fujiwara	2019	98	71	59.2	De novo AML	/	6
Grossmann	2021	1,000	66.8	/	No mentioned	FU 23.7 months	7
Hong	2020	125	66	53.6	AML-RGA, AML-MRC, AML-NOS, tAML	16 months/12 months	7
Hou	2015	500	51	57.0	De novo AML	FU 55 months	8
Inokuchi	2002	170	57	62.9	De novo AML	FU 20.6 months	6
Kihara	2014	197	<60 (89%)	/	De novo AML	FU 32.5 months	6
Middeke	2016	97	51	/	De novo AML, sAML	FU 67 months	8
Milosevic	2012	86	68	57.0	sAML	/	6
Montalban-Bravo	2020	415	70	64.0	AML-MRC	10.5 months/28.3 months	5
Morsia	2020	44	73.5	61.4	De novo AML, sAML, tAML	11 months/7 months	7
Mrózek	2019	136	53, 59	66.9	De novo AML	FU 81.6 months	7
Najima	2021	120	51	59.2	De novo AML, sAML, tAML	FU 60 months	6
Ni	2020	92	67	56.5	De novo AML (Elderly)	OS 9.7 months	5
		116	43	56.0	De novo AML (Younger <60 years)	/
Ohgami	2015	93	55	48.4	AML-MRC, AML-RGA	OS 221 days	5
Ok	2015	108	68	57.4	tMDS/AML	8 months/12.9 months	7
Österroos	2020	182	74	50.5	De novo AML, tAML	8.2 months/57 months	7
Rucker	2012	155	59	51.6	De novo AML, sAML, tAML	FU 36.6 months	8
Stahl	2018	655	65	58.2	De novo AML, sAML	OS 6.7 months	6
Stahl	2021	86	67	63.0	De novo AML, sAML, tAML	6.1m/12 months	6
Terada	2018	412	55.1	57.5	De novo AML	/	4
Thol	2018	96	50.1, 51.7	41.6	De novo AML, sAML	FU 74.4 months	7
Venton	2018	73	70	63.0	Post-MDN sAML	OS 4.7 months	5
Wang	2021	347	57	56.5	De novo AML	/	4
Wu	2021	81	54	54.0	No mentioned	OS 4 months	5
Yonada	2016	77	69, 70	62.3	AML-MRC, AML-RGA, tAML, AML-NOS	FU 48 months	6
Yu	2019	64	27.5	60.9	De novo AML	FU 23.5 months	6

AML, acute myeloid leukemia; AML-MRC, AML with myelodysplasia-related changes; AML-NOS, AML not otherwise specified; AML-RGA, AML with recurrent genetic abnormalities; FU, follow-up; M, male; MDS/AML, acute myeloid leukemia evolving from MDS; N, number; NOS, Newcastle-Ottawa Scale; OS, overall survival; Post-MDN sAML, acute myeloid leukemias secondary to myeloproliferative neoplasms (MPN); sAML, secondary acute myeloid leukemia; tAML, therapy-related acute myeloid leukemia; t-MDS, therapy-related myelodysplastic syndrome.

Prognostic Value of TP53 Mutations

Table 2 summarizes the TP53 mutational status across included studies. The frequency of TP53 mutation ranged from 2.6% (Ni et al. [34], younger AML cohorts) to 66.5% (Rucker et al. [38], AML cases with CK). Based on 11 studies with available data [16, 19, 23, 24, 25, 29, 30, 31, 32, 36, 38, 43], the median OS of AML patients with mutant TP53 ranged from 1.8 months to 10 months, while that of patients with wild-type TP53 ranged from 5.6 months to 35.6 months. The meta-analysis of data from 30 included studies (969 AML cases with TP53 mutation and 4,828 with TP53 wild-type) showed that TP53 mutation was associated with a statistically significant negative effect on the OS of AML patients with an HR of 2.40 (95% CI: 2.16–2.67, p < 0.05, Fig. 2) under a fixed-effect model due to nonsignificant heterogeneity (I²: 46.6%). As shown in Figure 3, patients with mutant TP53 also exhibited a worse prognosis than those with TP53 wild type in terms of RFS (HR: 2.40, 95% CI: 1.79–3.22), EFS (HR: 2.56, 95% CI: 1.97–3.31), and DFS (HR: 2.87, 95% CI: 1.88–4.38). A fixed-effect model was used owing to low heterogeneity between studies (I² <50%, Fig. 3).

Table 2.

Summary of TP53 status among included studies

Study ID	Applying NGS?	TP53 mutation		TP53 wild type		Frequency of TP53 mutation (%)
		n	median OS	n	median OS
Bories et al. [16] 2020	No	55	7.9 months	169	12.6 months	24.6
Chen et al. [17] 2021	Yes	15	/	189	/	7.4
Daher-Reyes et al. [18] 2021	Yes	25	/	153	/	14.0
Falk et al. [19] 2015	No	19	2 months	170	16 months	10.1
Fenwarth et al. [20] 2019	No	17	/	170	/	9.1
Fernandez-Pol et al. [21] 2017	No	12	/	55	/	17.9
Fujiwara et al. [22] 2019	No	12	/	86	/	12.2
Grossmann et al. [23] 2012	Yes	115	4.6 months	885	35.6 months	11.5
Hong et al. [24] 2020	Yes	15	10 months	110	18 months	12.0
Hou et al. [25] 2015	No	35	5 months	465	35 months	7.0
Inokuchi et al. [26] 2002	No	20	/	150	/	11.8
Kihara et al. [27] 2014	No	7	/	190	/	3.6
Middeke et al. [28] 2016	Yes	40	10% (3 years OS)	57	33.1% (3 years OS)	41.2
Milosevic et al. [29] 2012	No	14	1.8 months	59	5.6 months	19.2
Montalban-Bravo et al. [30] 2020	Yes	152	6.6 months	263	11.1 months	36.6
Morsia et al. [31] 2020	Yes	9	8 months	35	14 months	20.5
Mrózek et al. [32] 2019	Yes	69	4.8 months	67	7.2 months	50.7
Najima et al. [33] 2021	Yes	21	9.5% (2 years OS)	99	30.1% (2 years OS)	17.5
Ni et al. [34] 2020	Yes	13	/	79	/	14.1
	Yes	3	/	113	/	2.6
Ohgami et al. [35] 2015	Yes	17	/	76	/	18.3
Ok et al. [36] 2015	No	40	6.1 months	68	14.1 months	37.0
Österroos et al. [37] 2020	No	25	/	157	/	13.7
Rucker et al. [38] 2012	No	103	4.14 months	52	10.97 months	66.5
Stahl et al. [39] 2018	No	7	/	86	/	7.5
Stahl et al. [40] 2021	No	11	/	70	/	13.6
Terada et al. [41] 2018	No	31	14.5% (5 years OS)	381	33.7% (5 years OS)	7.5
Thol et al. [42] 2018	Yes	12	/	84	/	12.5
Venton et al. [43] 2018	Yes	20	4.4 months	36	6.5 months	35.7
Wang et al. [44] 2021	No	26	/	321	/	7.5
Wu et al. [45] 2021	Yes	10	/	71	/	12.3
Yonada et al. [46] 2016	No	12	/	65	/	15.6
Yu et al. [47] 2019	Yes	5	/	59	/	7.8

NGS, next-generation sequencing; OS, overall survival.

Fig. 2.

Forest plot of pooled HRs for OS comparing AML patients with TP53 mutation versus TP53 wild type. The HR with 95% CI exceeding 1 suggested a significantly higher risk of death in AML patients with TP53 mutation. HR, hazard ratio; CI, confidence interval; OS, overall survival.

Fig. 3.

Forest plot of pooled HRs for DFS, EFS, and RFS comparing AML patients with TP53 mutation versus TP53 wild type. HR, hazard ratio; CI, confidence interval; DFS, disease-free survival; EFS, event-free survival; RFS, relapse-free survival.

Egger's test revealed that potential bias might be potentially present in the analysis of OS (p = 0.02) and a trim-and-filled was used. Ten missing studies were filled by the trim-and-fill method, but the pooled estimate did not significantly change under the fixed-effect model (before: LogHR 0.874, 95% CI: 0.768–0.981; after filled analysis: LogHR 0.748, 95% CI: 0.648–0.847), indicating that the result remained robust albeit with the presence of potential bias.

Additional Subgroup Analysis

As displayed in Figure 4a, subgroup analyses stratified by intensive treatment suggested that TP53 mutation was associated with a more negative OS in intensively treated patients (HR: 2.73, 95% CI: 2.39–3.13), compared with that in patients treated by nonintensive therapy (HR: 1.89, 95% CI: 1.58–2.26) (test of interaction: p = 0.001). We further analyzed each subgroup based on the median age (≥65 years old or <65 years old) of enrolled subjects. In the “intensively treated” group, no significant difference was found between median age <65 and ≥65 years groups (test of interaction: p = 0.83), indicating that intensively treated AML cases with mutant TP53 might demonstrate a worse OS regardless of age (Fig. 4b). Of note, the median age of the “nonintensively treated” group was more than 65; therefore, among AML subjects with median age ≥65, the nonintensive treatment showed a better OS than the intensive treatment (HR: 1.89 vs. 2.98, test of interaction: p < 0.05; Fig. 4).

Fig. 4.

Subgroup analysis of pooled HRs for OS according to intensive treatment in total AML patients (a) and further analysis of pooled HRs for OS according to the median age of 65 years old in intensively treated AML patients (b).

Relations between TP53 Mutations and Cytogenetics

Cytogenetics information was frequently used for AML risk stratification, and we conducted additional analyses to investigate the relation between adverse cytogenetics and TP53 mutations. Pooled by 12 studies, results suggested that there was a significant difference in OS between patients with adverse cytogenetics and patients with favorable or intermediate cytogenetics (HR: 2.03, 95% CI: 1.74–2.37, I²: 32.4%). Furthermore, compared to patients with TP53 wild type, patients harboring TP53 mutation were at an elevated risk of adverse cytogenetics (OR: 23.22, 95% CI: 13.84–38.95, I²: 18.6%). It is indicated that TP53 mutation was significantly correlated with adverse cytogenetics, which might predict a worse OS. Forest plot is shown in Figure 5.

Fig. 5.

Forest plot of pooled HRs for OS comparing AML patients with adverse cytogenetics versus nonadverse cytogenetics (a) and the relation between TP53 mutation and adverse cytogenetics (b). HR, hazard ratio; CI, confidence interval; OR, odds ratio; OS, overall survival.

Discussion

In this meta-analysis based on 32 eligible studies involving 7,062 adult AML cases, results demonstrated that patients with TP53 mutations exhibited a significantly worse survival than those with TP53 wild type, given the significant HRs greater than 1 for OS, EFS, DFS, or RFS, which validated the prognostic potential of TP53 mutation in patients with AML in a comprehensive manner.

AML is characterized by a spectrum of gene mutations, including TP53, that stimulate the self-renewal of leukemia stem cells or leukemia-initiating cells (LICs) with blockade of differentiation [48]. TP53 mutations usually result in loss of function of the p53 protein, which demonstrates pivotal functions in normal hematopoietic stem cells (HSC) being involved in their proliferation, differentiation, and apoptosis [6]. Wild-type p53 maintains HSC quiescence and negatively regulates HSC aberrant self-renewal [49]. Zhao et al. [49] demonstrated based on a combined mosaic mouse model that mutant p53, in concert with oncogenic Kras, induced aberrant HSC self-renewal and repopulated leukemic blasts, paving a way for the formation of LICs and propagating AML. Also, another mouse model showed that mutant p53 synergizing with FLT3-ITD enhanced the self-renewal potential of LICs, which contributed to leukemia development [50]. Given the critical role of p53 in the initiation, development, and progression of AML, TP53 mutations may constitute a useful tool for prognostication of AML patients, with its prognostic potential systematically validated by this meta-analysis using a collection of previously published data.

Development of AML represents a variety of cytogenetic and molecular abnormalities. Cytogenetics at diagnosis serves as the single most important prognostic indicator for AML and patients with adverse cytogenetics were reported to be less likely to achieve a complete response and long-term survival [46]. However, almost half of AML patients lack cytogenetic abnormality, thus being allocated to the intermediate-risk group with significant clinical heterogeneity [17]. Also, there exists a time gap between the initiation of induction therapy and the obtainment of available cytogenetic information [46]. Thus, to improve the therapeutic outcomes and prognosis, it is of importance to establish a prognostic stratification that takes into consideration a combination of cytogenetic abnormality with gene aberrations, such as TP53 mutation, which have been proved to correlate with CK [38]. The present analysis showed an elevated risk of unfavorable cytogenetics in favor of patients with TP53 mutation and a close association between adverse cytogenetics and poor prognosis, supporting the role of TP53 mutation in prognostic stratification and in assisting the risk stratification scheme based on cytogenetic abnormality.

It has been reported that p53 loss of function correlates with chemotherapy resistance [38]. Results of our subgroup analyses are consistent with previous evidence. TP53 mutations were more significantly associated with a poor OS in adult AML patients treated with intensive chemotherapy treatment, as compared to those treated with nonintensive treatment. Similar results were observed when subgroups were stratified according to the median age of 65. The aforementioned results indicated the nonuse of intensive therapies for newly diagnosed AML patients harboring TP53 mutations. Notably enough, treatment options beyond standard intensive therapies for these patients represent a conundrum. Nonintensive therapies with hypomethylating agents, including azacitidine and decitabine, are frequently provided, especially to elderly AML, yet their curative effects are limited [51]. APR-246 is a novel small molecule designed to shift mutant p53 toward the wild-type conformation and induce apoptosis of TP53 mutated cancer cells. It has been reported that the combination of APR-246 and azacitidine in AML patients with mutant TP53 showed encouraging response rate with good tolerability [52]; however, the current evidence has yet reached the sufficient level and additional research aimed at therapies targeting TP53 mutations in AML patients is needed.

TP53 mutations have been evaluated as a dichotomous variable (mutant vs. wild type) in most of the studies performed so far, as shown in this meta-analysis. More attention should be paid to other novel aspects of TP53 mutations, like functional classifications and variant allele frequencies (VAF), in future studies. There are distinct types of TP53 mutations – missense, nonsense, insertions, deletions, and splice-site mutations. However, it is not yet elucidated whether these types would exert a uniformly poor outcome in AML. Different methods have been applied for the functional classification of TP53 mutations. Of them, relative fitness score, representing an algorithm based on the average of the relative enhancement or depletion of a specific TP53 mutation detected at three time points, has been suggested to hold prognostic significance and awaits further investigation [4, 53]. Besides, Prochazka et al. categorized TP53 mutations according to their VAF into groups of >40% (major AML clone), 20–40%, and <20% (subclones), suggesting that a dismal outcome also occurred in leukemic subclones with a VAF <20% [54]. Thus, more robust evidence for the prognostic role of subclonal TP53 mutations is encouraged.

Based on the close relation between TP53 mutations and an adverse prognosis, their utility in epidemiology and diagnostic workup should be noted. Somatic TP53 mutations observed at diagnosis of AML have been well characterized, yet the presence of detectable mutations years before diagnosis should also be attached importance. Desai and colleagues [55] reported that all healthy women with mutations in TP53 at baseline were eventually developed into AML during a median follow-up of 9.6 years. It is reasonable for genetic screening for mutations in TP53 to be utilized in the setting of AML surveillance.

Several limitations should be seen in our analysis. First, there existed a mixture of prospective and retrospective studies, involving different types of AML in varied centers, which could be an important source of heterogeneity. Subgroup analyses should be treated with caution as the grouping was made based on the available information provided by the included studies where heterogeneity was also inherent regarding the actual treatment regimens. Also, there are limited data in included studies for us to investigate the prognostic value of TP53 mutations based on different VAFs or distinct types. Another limitation included the presence of publication bias, which might be attributed to the fact that we only included studies published in English.

Conclusion

In conclusion, this present meta-analysis investigated the prognostic value of TP53 mutations in adult AML in a comprehensive manner, demonstrating that TP53 mutations were significantly associated with an unfavorable prognosis and could be incorporated into the prognostication and decision-making in the management of AML. Future studies are encouraged to investigate the novel aspects of TP53 mutations in terms of VAF and functional classification and to explore the therapies targeting TP53 mutations in the setting of AML.

Statement of Ethics

An ethics statement is not applicable because this study is based exclusively on the published literature.

Conflict of Interest Statement

The authors have no conflicts of interest to declare.

Funding Sources

This study did not receive any funding.

Author Contributions

Conception and design, administrative support, provision of study materials, collection and assembly of data, data analysis and interpretation, manuscript writing, and final approval of manuscript: all authors.

Data Availability Statement

All data generated or analyzed during this study are included in this article and its online supplementary material. Further enquiries can be directed to the corresponding author.

Supplementary Material

Supplementary data

Funding Statement

This study did not receive any funding.