TP53 Mutations in AML Patients Are Associated with Dismal Clinical Outcome Irrespective of Frontline Induction Regimen and Allogeneic Hematopoietic Cell Transplantation

Simple Summary TP53 mutations are adverse-risk genetic aberrations in acute myeloid leukemia (AML). The optimal treatment approach in patients with TP53-mutated (TP53MUT) AML remains unclear. We aimed to evaluate the prognostic implications of different frontline treatment strategies and transplantation for patients with TP53MUT AML. Patients treated with intensive induction or azacitidine-venetoclax induction had no significant improvement in survival compared to patients treated with other HMA regimens despite having higher complete remission rates. Transplantation was not significantly associated with improved outcomes in time-dependent or landmark analysis, however, transplanted patients with lower TP53MUT variant allele frequency (VAF) at the time of diagnosis had superior outcomes compared to transplanted patients with higher TP53 VAF. Current therapeutic strategies remain ineffective for TP53MUT AML patients, which highlights the urgent need for new treatment strategies for this high-risk population. Abstract TP53 mutations are associated with extremely poor outcomes in acute myeloid leukemia (AML). The outcomes of patients with TP53-mutated (TP53MUT) AML after different frontline treatment modalities are not well established. Allogeneic hematopoietic cell transplantation (allo-HCT) is a potentially curative procedure for AML; however, long-term outcomes among patients with TP53MUT AML after allo-HCT are dismal, and the benefit of allo-HCT remains controversial. We sought to evaluate the outcomes of patients with TP53MUT AML after treatment with different frontline induction therapies and allo-HCT. A total of 113 patients with TP53MUT AML were retrospectively evaluated. Patients with TP53MUT AML who received intensive or azacitidine-venetoclax induction had higher complete remission rates compared to patients treated with other hypomethylating-agent-based induction regimens. However, OS and EFS were not significantly different among the induction regimen groups. Allo-HCT was associated with improved OS and EFS among patients with TP53MUT AML; however, allo-HCT was not significantly associated with improved OS or EFS in time-dependent or landmark analysis. While the outcomes of all patients were generally poor irrespective of therapeutic strategy, transplanted patients with lower TP53MUT variant allele frequency (VAF) at the time of diagnosis had superior outcomes compared to transplanted patients with higher TP53 VAF. Our study provides further evidence that the current standards of care for AML confer limited therapeutic benefit to patients with TP53 mutations.


Introduction
The TP53 gene, located on chromosome 17p, plays a central role in anti-tumour response by regulating the cell cycle and apoptosis [1]. Mutations in TP53 occur in approximately 10-20% of acute myeloid leukemia (AML) patients and are associated with an extremely poor prognosis [2]. TP53 mutation is an adverse-risk genetic abnormality according to the 2022 ELN risk classification of AML due to its association with complex karyotype, advanced age, and resistance to standard therapies [3]. TP53 mutation defines a distinct standalone entity within the International Consensus Classification of Myeloid Neoplasms and Acute Leukemia [4].
In fit patients with high-risk AML, the standard treatment approach is induction therapy, followed by allogeneic hematopoietic cell transplantation (allo-HCT) as a curative modality for those in remission [3]. Patients with TP53-mutated (TP53 MUT ) AML have a lower probability of achieving remission when treated with intensive chemotherapy and, consequently, have poor outcomes; for those who achieve remission, allo-HCT provides a relatively small improvement in long-term survival [2,[5][6][7][8][9].
Given that there is limited data comparing the outcomes after different treatment modalities for patients with TP53 MUT AML, we retrospectively reviewed our institution's experience with TP53 MUT AML. We assessed the prognostic implications of different frontline treatment strategies and transplantation for patients with TP53 MUT AML.

Patients and Treatment
We screened all adult University Health Network (UHN) patients who were diagnosed with AML with blasts ≥20% from February 2015 to December 2021 and who had molecular testing results available. Patients' medical records were retrospectively assessed with respect to demographic and clinicopathological data, treatment modalities, and survival outcomes. The study was approved by the University Health Network Research Ethics Board.
According to the American Society of Transplantation and Cellular Therapy recommendations for unfavourable-risk AML, allo-HCT was offered to patients with TP53 MUT AML in CR if a suitable donor was identified [10,11]. Patients received 1-2 cycles of consolidation chemotherapy before proceeding to treatment with allo-HCT.

Mutational Analysis and Determination of TP53 Allelic Status
Next-generation sequencing (NGS) was performed using a custom gene panel of 49 myeloid genes (Oxford Gene Technologies, Oxford, UK) or 54 myeloid genes (Illumina, San Diego, CA, USA) and run on the MiSeq platform (Illumina), as previously described [12][13][14]. The limit of detection for variant calling was 2%. For both panels, for 13/41 genes, the complete coding regions were analyzed, and for 28/41 genes, the same exonic hotspots were analyzed (Tables S1 and S2). All consensus-coding TP53 exons were interrogated. Interpretation and classification of variants was performed as Cancers 2023, 15, 3210 3 of 12 previously described [12]. Briefly, pathogenic (AMP/ASCO Tier 1) variants consisted of (a) loss-of-function truncating variants (nonsense, frameshift, and splicing) and (b) missense or in-frame variants (typically, within the DNA-binding domain) demonstrating loss-of-function (e.g., p.R110P) [15,16]. Likely pathogenic (AMP/ASCO Tier 2) variants consisted of (a) missense or in-frame variants involving the same residue (e.g., R110) as an established pathogenic variant and (b) missense or in-frame variants (typically, within the DNA binding domain) classified as disruptive in the TP53 IARC database [17]. Variants of uncertain significance (VUS, AMP/ASCO Tier 3) included variants classified as partially or fully transcriptionally active according to the TP53 IARC database as well as variants with no functional evidence of deleterious potential. VUS and benign polymorphisms were excluded from analysis. When patients had multiple mutations in the same gene, the higher variant allele frequency (VAF) was used for analysis. TP53 mutations were visualized using cBioPortal [18,19].
Patients were considered to be double/multi-hit TP53 when (A) at least two TP53 variants were detected via NGS, (B) one TP53 variant detected via NGS co-occurred with cytogenetic loss of TP53, or(C) one TP53 variant was detected via NGS with a VAF of ≥55%, as previously described [20].

Karyotype Analysis
In accordance with the International System for Human Cytogenetic Nomenclature's guidelines, karyotypes were obtained from diagnostic bone marrow samples and described as appropriate. The cytogenetic loss of TP53 was determined as previously described [21].

Statistical Analysis
Categorical variables were summarized as counts and percentages. Continuous variables were summarized as medians and ranges. Evaluation of patient outcome including overall survival (OS) and event-free survival (EFS) was carried out through retrospective analysis of patient records. OS was calculated from the date of diagnosis to last follow-up or death. EFS was calculated from the date of diagnosis to last follow-up, relapse, or death. Log-rank test was used to compare Kaplan-Meier survival curves. Univariate Cox proportional-hazard regression models were fitted to identify prognostic factors for OS and EFS. Cox proportional-hazard regression was used to evaluate the prognostic impact of transplantation as a time-dependent variable, as transplant (if any) can occur during the follow-up period. Differences in categorical variables were assessed using the Fisher's exact test. Differences in numerical variables across different groups were assessed using the Kruskal-Wallis and Wilcoxon rank-sum tests. Statistical analysis was performed using R version 4.0.5 (R Core Team (2020). R: A language and environment for statistical computing, Vienna, Austria).
Given the dismal outcomes of patients with TP53 MUT AML treated with allo-HCT, we sought to identify factors that may predict favourable outcomes with respect to transplantation. When using the median TP53 VAF of transplanted patients (45%) as the cut-off, transplanted patients with low TP53 VAF at the time of diagnosis had superior OS To further confirm that the favourable outcome associated with transplantation may be confined to patients with low TP53 mutational burden, we performed subgroup landmark analyses for treated patients stratified by the median TP53 VAF of the transplanted To further confirm that the favourable outcome associated with transplantation may be confined to patients with low TP53 mutational burden, we performed subgroup landmark analyses for treated patients stratified by the median TP53 VAF of the transplanted patients (45%). For patients with low TP53 VAF, landmark analysis for OS that excluded patients who died or were lost to follow-up before 4.11 months following diagnosis (corresponding to the median time from diagnosis to allo-HCT) indicated that allo-HCT significantly improved OS (HR: 0.21, 95% CI: 0.06-0.73, p = 0.014) ( Figure S3A). Similarly, landmark analysis for EFS that excluded patients who died, were lost to follow-up, or relapsed before 4.11 months following diagnosis indicated that allo-HCT marginally improved EFS (HR: 0.36, 95% CI: 0.13-1.01, p = 0.053) among patients with low TP53 VAF ( Figure S3B). In contrast, for patients with high TP53 VAF, landmark analysis for OS that excluded patients who died or were lost to follow-up before 4.11 months following diagnosis indicated that allo-HCT did not significantly improve OS (HR: 1.10, 95% CI: 0.44-2.71, p = 0.84) ( Figure S3C). Similarly, landmark analysis for EFS that excluded patients who died, were lost to follow-up, or relapsed before 4.11 months following diagnosis indicated that allo-HCT did not significantly improve EFS (HR: 1.06, 95% CI: 0.41-2.71, p = 0.91) among patients with high TP53 VAF ( Figure S3D).

Discussion
In this study, we report a single-institution evaluation of different treatment modalities for patients with TP53 MUT AML. While the use of any treatment was associated with better outcomes than best supportive care, intensive induction chemotherapy was not associated with a better outcome than HMA therapy. Furthermore, while transplanted patients had significantly superior outcomes compared to non-transplanted patients, this association was lost in time-dependent analysis and in landmark analysis.
Anti-cancer responses to cytotoxic chemotherapies are critically dependent on wildtype TP53, as the activation of p53 by chemotherapy-induced DNA damage induces apoptosis and cancer cell death [22,23]. As such, standard induction treatments using DNAdamaging drugs may be a suboptimal therapeutic approach for patients with TP53 MUT AML. Prior clinical studies have demonstrated that patients with TP53 MUT AML have poor outcomes and response rates when treated with intensive chemotherapy regimens [24][25][26][27][28][29]. In addition, other less-intensive therapies such as hypomethylating agents and venetoclax have shown varying levels of efficacy among TP53 MUT AML patients [30][31][32][33]. In our cohort, patients with TP53 MUT AML had no significant differences in outcomes after treatment with intensive and non-intensive frontline regimens despite the younger age of the intensively treated patients. While the addition of venetoclax significantly improved complete remission rates among patients with TP53 MUT AML treated with HMA-based regimens, survival was not significantly improved with venetoclax, which is in line with prior reports [34][35][36]. Indeed, TP53 mutations have been shown to confer resistance to BCL-2 inhibition in preclinical studies [37][38][39]. Clinical trials for novel targeted agents and immunotherapeutic approaches such as TP53 MUT re-activators and anti-CD47 blockade have shown potential in the frontline setting of TP53 MUT AML and should be explored further in this high-risk population [40].
Allo-HCT is considered to be a potentially curative option for high-risk AML patients; however, whether transplantation confers a beneficial impact among patients with TP53 MUT AML is controversial. Prior studies have largely supported the use of allo-HCT in these patients as they have reported superior outcomes in transplanted patients compared to non-transplanted patients [27,35,41,42].Conversely, there is evidence in the literature that patients with TP53 MUT AML do not significantly benefit from allo-HCT [34]. We previously evaluated the impact of allo-HCT among patients with AML with myelodysplasia-related changes and reported that allo-HCT had a significant benefit among TP53 MUT patients [43]. However, consistent with our data presented herein, TP53 MUT patients did not experience sustained and plateauing long-term outcomes and survival rates after allo-HCT. Since the outcomes of transplanted and non-transplanted patients may be confounded due to guarantee-time bias, we performed a prognostic evaluation of allo-HCT using timedependent and landmark analyses [44]. In these analyses, TP53 MUT patients who received allo-HCT did not have significantly superior outcomes compared to non-transplant patients, suggesting that the differences in outcome according to allo-HCT status are confounded by guarantee-time bias. Despite this, among transplanted patients, low TP53 VAF at the time of diagnosis was associated with superior outcomes. Further landmark analysis of patient subgroups revealed that the favourable outcome associated with transplant is confined to patients with low TP53 VAF, who had significantly improved outcomes with allo-HCT compared to non-transplanted patients with low TP53 VAF. Our findings are consistent with previous reports demonstrating associations between low TP53 VAF and superior outcomes in transplanted patients with TP53 MUT AML [35,45]. As such, TP53 VAF may have clinical utility for guiding the decision-making process of whether TP53 MUT patients should undergo transplantation. While other studies have shown an adverse prognostic impact of cytogenetic loss of TP53 among transplanted patients with TP53 MUT AML, we did not observe a significant difference in our cohort [46]. Future studies are needed to optimize transplant approaches and post-transplant monitoring practices and to explore the prognostic benefit of allo-HCT for TP53 MUT AML patients.
The updated 2022 ELN guidelines for the management of AML recommend for TP53 mutational analysis to be performed within the first cycle of chemotherapy [3]. At our institution, TP53 sequencing is performed via NGS, which has a relatively longer turnaround time compared to other diagnostic tests. The lack of benefit for intensive chemotherapy among patients with TP53 MUT AML suggests that genetic results for TP53 should be obtained before beginning treatment to avoid ineffective and aggressive therapy. Rapid and cost-effective methods such as immunohistochemistry have shown potential for predicting TP53 mutational status and should be further explored in the clinical setting [47,48].
We acknowledge that our study has several limitations. First, our study is retrospective in nature and includes limited patient numbers in its subgroup analyses. Future multicenter prospective studies are needed to confirm our data. Second, transplanted patient outcomes may be influenced by factors such as patient performance status, conditioning regimen intensity, and post-transplant graft-versus-host disease. Further studies are needed to evaluate the impact of these factors on patient outcomes and to determine the optimal allo-HCT approach for patients with TP53 MUT AML. Third, this study only focused on patients with TP53 MUT AML; future prospective studies should include patients with and without TP53 mutations to evaluate the clinical impact of TP53 mutations across different treatment groups.

Conclusions
In conclusion, outcomes among patients with TP53 MUT AML after intensive therapy and allo-HCT remain poor; however, patients with low TP53 VAF at the time of diagnosis may still benefit from transplantation. Our study provides further evidence that patients with TP53 MUT AML derive limited benefit from current treatments and highlights the urgent need for new treatment strategies.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cancers15123210/s1, Figure S1: Outcomes according to presence/absence of TP53 mutations in the DNA binding domain. Figure S2: Outcomes according to transplant status. Figure S3: Outcomes according to TP53 VAF and transplant status; Table S1: Gene panel for targeted sequencing; Table S2: Exon coverage for hotspot genes; Table S3: Summary of additional clinicopathological features.