Prognostic impact of ‘multi-hit’ versus ‘single-hit’ TP53 alteration in patients with acute myeloid leukemia: results from the Consortium on Myeloid Malignancies and Neoplastic Diseases

Abstract

While there is clear evidence to suggest poorer outcome associated with multi-hit (MH) TP53 mutation (TP53MT) compared to a single-hit (SH) mutation in lower-risk myelodysplastic syndrome (MDS), data are conflicting in both higher-risk MDS and acute myeloid leukemia (AML). We conducted an in-depth analysis utilizing data from ten US academic institutions to study differences in molecular characteristics and outcomes of SH (N=139) versus MH (N=243) TP53MT AML. Complex cytogenetics were more common in MH than in SH TP53MT AML (P<0.001); whereas ASXL1 (P<0.001), RAS (P<0.001), splicing factor (P=0.003), IDH1/2 (P=0.001), FLT3 ITD (P<0.001) and NPM1 (P=0.005) mutations clustered significantly with SH TP53MT AML. Survival after excluding patients who received best supportive care alone was dismal but not significantly different between patients with SH or MH disease (event-free survival: 3.0 vs. 2.20 months, respectively, P=0.22; overall survival: 8.50 vs. 7.53 months, respectively, P=0.13). In multivariable analysis, IDH1 mutation and allogeneic hematopoietic stem cell transplantation as a time-dependent covariate were associated with superior event-free survival (hazard ratio [HR]=0.44, 95% confidence interval [95% CI]: 0.19-1.01, P=0.05 and HR=0.34, 95% CI: 0.18-0.62, P<0.001) and overall survival (HR=0.24, 95% CI: 0.08-0.71, P=0.01 and HR=0.28, 95% CI: 0.16-0.47, P<0.001). Complex cytogenetics (HR=1.56, 95% CI: 1.01-2.40, P=0.04) retained an unfavorable significance for overall survival. Our analysis suggests that MH TP53MT is less relevant in independently predicting outcomes in patients with AML than in those with MDS.

Introduction

TP53 is the most frequently mutated gene across all malignancies and is associated with a poor prognosis across many cancer types with suboptimal responses to standard-of-care therapies.^1,2 TP53-mutated (TP53^MT) acute myeloid leukemia (AML) is strongly associated with large structural and complex cytogenetic abnormalities, often seen among recipients of prior cytotoxic therapies.^3-8 Despite the increasing availability of novel therapies, the median overall survival (OS) of patients with TP53^MT AML remains in the range of 6-9 months, irrespective of therapy intensity.^9-13 Single-hit TP53^MT is associated with clonal hematopoiesis and may not be directly leukemogenic unless accompanied by subsequent hits that could be secondary to cytotoxic stress.^4,14,15 There are conflicting reports regarding the prognostic impact of allelic state, specifically bi-allelic alteration/“multi-hit” (MH) TP53^MT versus mono-allelic/ “single-hit” (SH) TP53^MT among patients with myeloid neoplasms.^10,16-18 Bernard et al. performed extended genetic profiling in a large cohort of patients with myelodysplastic syndrome (MDS) and showed that not all TP53^MT have equivalent impact on survival.¹⁹ Patients with MDS harboring SH TP53^MT had similar outcomes to their counterparts with TP53 wild-type disease. Conversely, multiple hits caused by either multiple mutations of TP53/copy-neutral loss of heterozygosity or mono-allelic TP53^MT with deletion of the other TP53 allele were associated with inferior clinical outcomes. However, in patients with high-risk MDS with excess blasts or AML, it was recently demonstrated that TP53 allelic state (SH vs. MH TP53^MT) did not predict differences in clinical outcome.¹⁰ The authors concluded that further risk stratification by TP53 allelic state may be less relevant among patients with advanced MDS or AML. Here, we present real world data on a large cohort of patients with TP53^MT AML and report their clinical characteristics, therapy received, and outcome based on TP53 allelic state.

Methods

We conducted a retrospective study through the Consortium on Myeloid Malignancies and Neoplastic Diseases (COMMAND) consortium (a collaboration of acute leukemia experts from 10 US academic institutions) to analyze the prognostic impact of MH versus SH TP53^MT on outcomes of adult (≥18 years) patients with AML. A total of 382 adults with TP53^MT (139 SH, 243 MH) AML who were diagnosed between November 2012 and May 2023 were evaluated, and their baseline characteristics, molecular profile, and treatment outcomes were compared based on SH versus MH TP53^MT status. The current cohort of 382 patients was increased from the 291 patients who were included in our previous publication;⁹ furthermore, the current cohort has more robust TP53 gene annotation data and longer follow-up. These features of the cohort reported here enable a more comprehensive evaluation regarding the impact of TP53 mutation burden on clinical outcome.

MH TP53^MT was defined by the presence of two or more distinct TP53^MT regardless of variant allele frequency (VAF) or a single TP53^MT associated with (i) cytogenetic abnormalities involving chromosome 17p (e.g. abnormality of 17p or monosomy 17) or (ii) a VAF of ≥55%, as previously reported by Grob et al.¹⁰ Loss of heterozygosity was not assessed in all patients in this dataset.

AML was diagnosed as per the 2016 World Health Organization classification.²⁰ Response to treatment was defined according to 2017 European LeukemiaNet consensus guidelines.²¹ Next-generation sequencing was performed at diagnosis using DNA extracted from bone marrow aspirate specimens with post-sequencing analysis of tumor-associated mutations. Next-generation sequencing testing was developed, and its performance characteristics determined by the participating institutions in compliance with Clinical Laboratory Improvement Amendments requirements. The next-generation sequencing panel had a sensitivity of ≥5% VAF with a minimum depth coverage of 250x.

The study was conducted after obtaining approval from the Institutional Review Board, adhering to the ethical standards of the Declaration of Helsinki of 1975, as revised in 2000.

Statistical analysis

Continuous variables are summarized as the median (range) while categorical variables are reported as frequency (percentage). Duration of response (complete [CR] or complete with incomplete blood count recovery [CRi]) was defined from the time of onset of response to progression or death due to any reason, whichever occurred earlier. The Kaplan-Meier method was used to estimate event-free survival (EFS), defined as time from diagnosis to relapse or death. The median overall survival (OS) was calculated from time of diagnosis to death or last follow-up. Cox proportional hazards regression models were used to determine the univariate and multivariate predictors of overall mortality and progression. Allogeneic hematopoietic stem cell transplantation (HSCT) was treated as a time-dependent covariate. Multivariable models included all significant univariate predictors. All tests were two-sided with a P value <0.05 considered statistically significant.

Results

Baseline characteristics

A total of 382 adult patients with TP53^MT AML (SH, 139; MH, 243) were identified. Among the 243 patients with MH TP53^MT, 57 patients had multiple TP53^MT, 58 patients had TP53^MT with VAF of ≥55%, and 128 patients had single TP53^MT associated with cytogenetic abnormalities involving chromosome 17p (e.g., abnormality of 17p or monosomy 17). The median age was 67 (range, 23-90) and 66.5 (range, 18-97) years in the SH and MH TP53^MT AML groups, respectively (P=0.86) (Table 1). Thirty-nine (33%) and 70 (29%) patients had secondary AML in the SH and MH groups, respectively (P=0.34). Among these 109 patients with secondary AML, 11 (10%) patients had JAK2-mutated myeloproliferative neoplasm in blast phase, of whom four (3%) were in the SH group and seven (3%) in the MH group (P=0.82). The median TP53^MT VAF was 22% (range, 5-49%) and 50% (range, 5-98%) in the SH and MH TP53^MT AML groups (P<0.001), respectively. The proportion of patients with complex cytogenetics was higher in the MH group than in the SH group (93% vs. 58%, P<0.001). In subgroup analysis, we looked at baseline characteristics of patients with IDH1 or IDH2 co-mutated AML. Patients with secondary AML had a higher proportion of IDH2-mutated disease than IDH1-mutated disease (46% vs. 27%, P=0.15) and complex cytogenetics (54% vs. 27%, P=0.24), but these differences were not statistically significant (Online Supplementary Table S1).

Table 1.Baseline characteristics, treatment, and outcome in single-hit and multi-hit TP53.

Molecular profile and somatic co-mutation pattern

An overview of TP53 domains, distribution of TP53 variants and position on the TP53 protein are illustrated in Figure 1. The occurrences of somatic co-mutations were comparable between the SH (67%) and MH (60%) groups (P=0.22). ASXL1 (16% vs. 7%, P<0.001), RAS (15% vs. 6%, P<0.001), splicing factor (12% vs. 4%, P=0.003), IDH1/2 (11% vs. 4%, P=0.001), FLT3-ITD (11% vs. 2%, P<0.001) and NPM1 (6% vs. 1%, P=0.005) mutations were more frequent in the SH group than in the MH group. The somatic co-mutation patterns and frequency of co-mutations in the SH and MH groups, are illustrated in Figure 2 and Online Supplementary Figure S1, respectively. Eleven (46%) patients had IDH1, and 13 (54%) patients had IDH2 mutations. Two (18%) and seven (53%) patients with IDH1 and IDH2 mutations, respectively, had MH TP53^MT. There were no differences in the co-mutational patterns among patients with IDH1/IDH2 co-mutated disease with the lone exception of JAK2 mutations, which were more common in the IDH2 co-mutated group (38.5% vs. 0%, P=0.04) (Online Supplementary Table S1).

Figure 1.Overview of TP53 domains, structures, and distribution of TP53 variants detected, positioned on the TP53 protein. Variants from patients with mono-allelic TP53 are depicted at the top and those from patients with multiple TP53 hits at the bottom. Missense mutations are shown as gold circles and all other variants, including truncated mutations corresponding to splice site variants, nonsense, nonstop, and frameshift deletions or insertions, are shown as blue circles. NTD: N-terminal domain; TAD: transactivation domain; PRD: proline-rich domain; DBD: DNA-binding domain; TD: tetramerization domain; BD: basic domain; CTD: C-terminal domain.

Treatment and outcome

A significantly higher proportion of patients in the MH group received hypomethylating agents plus venetoclax compared to the SH group (29% vs. 19%, P=0.01). However, the proportion of patients who received intensive chemotherapy, hypomethylating agent-based therapy or other low-intensity chemotherapy (low-dose cytarabine, IDH2 inhibitor alone or an investigational agent) were comparable between the two groups (Table 1). The response rates (CR/ CRi) were comparable between the SH and MH groups (28% vs. 22%, P=0.21). Among the 91 (26%) patients with CR/CRi, 28 (31%) were did not have measurable residual disease (MRD), assessed by flow cytometry, after induction. The MRD-negative CR rates with intensive versus non-intensive chemotherapy were not significantly different (10% vs. 7%, respectively, P=0.78). Similarly, a comparable proportion of patients underwent allogeneic HSCT after induction (12% vs. 14%, P=0.53). In subgroup analysis, there was a significant difference in response rate between patients with IDH1 co-mutated disease (54.4%) and those with IDH2 co-mutated disease (0%) (P=0.003) (Online Supplementary Table S1). The median duration of response was 7.77 versus 12.83 months in the SH and MH groups, respectively (P=0.73) (Figure 3A).

Predictors of response

Predictors of response (CR/CRi) to induction chemotherapy were evaluated, and the results are summarized in Online Supplementary Table S2. The co-occurrence of RAS (NRAS or KRAS) (P=0.02) and IDH2 mutations (P=0.03) had a negative impact on response rate. Conversely, the co-occurrence of IDH1 mutation (P=0.02) and induction with a hypomethylating agent plus venetoclax (P<0.001) was associated with better responses. In this cohort of adverse-risk TP53^MT AML, age ≥65 years (P>0.99), secondary AML (P=0.58), therapy-related AML (P>0.99), and complex cytogenetics (P>0.99) did not have significant impacts on achieving response.

Event-free survival

Considering the significantly higher proportion of patients receiving supportive care alone in the SH group compared to the MH group, we excluded these patients from the survival analysis. The median EFS in months was not significantly different between the SH and MH groups (3.0 and. 2.20, respectively, P=0.22) (Figure 3B). However, there was a statistically significant difference in EFS between the SH and MH groups (3.0 vs. 2.13, P=0.02), utilizing the definition of MH as per the International Consensus Classification, i.e., two distinct TP53^MT with VAF >10% or a single TP53^MT with (i) 17p deletion; (ii) VAF of >50%; or (iii) copy-neutral loss of heterozygosity at the 17p TP53 locus.²² In univariate analysis for EFS (Online Supplementary Table S3), complex cytogenetics adversely affected outcome (P=0.04). In contrast, ASXL1 mutation (P=0.02), IDH1 mutation (P=0.01), hypomethylating agent plus venetoclax induction (P<0.001), and allogeneic HSCT as a time-dependent covariate (P<0.002) were associated with favorable EFS in univariate analysis. In multivariable analysis for EFS, IDH1 co-mutation (HR=0.44, 95% CI: 0.19-1.01, P=0.05), hypomethylating agent plus venetoclax induction (HR=0.53, 95% CI: 0.41-0.70, P<0.001) and allogeneic HSCT (HR=0.34, 95% CI: 0.18-0.62, P<0.001) retained statistically significant associations with favorable outcomes.

Overall survival

After excluding patients who received supportive care alone, we calculated the median OS. The median OS in months was not significantly different between the SH and MH groups (8.50 vs. 7.53, respectively, P=0.13) (Figure 3C). Likewise, we did not observe a significant difference in OS between the SH and MH groups (8.0 vs. 8.0, respectively, P=0.32), utilizing the MH definition as per the International Consensus Classification. We looked at the impact of complex cytogenetics on OS in the SH and MH groups and found that OS was better in the SH and MH subgroups without complex cytogenetics (9.97 and 10.07 months, respectively) than in those with complex cytogenetics (6.2 and 7.13 months, respectively) (P=0.008) (Figure 3D). We performed landmark analysis from the time of achievement of CR/CRi until last follow-up or death; allogeneic HSCT recipients had better OS compared to non-allogeneic HSCT recipients in both the SH group (not reached vs. 9.63 months, respectively) and the MH group (24.3 vs. 9.6 months, respectively) (P=0.001) (Figure 3E). In another subset analysis among transplanted patients, those who were transplanted in MRD-negative CR (N=12) had a longer median OS compared to those with MRD-positive disease (N=43) (46.1 vs. 25.47 months, respectively, P=0.15) although the difference was not statistically significant probably due to a smaller sample size. We performed a similar analysis to look at OS in relation to complex cytogenetics and TP53 allelic state among patients who achieved CR/CRi and did or did not receive allogeneic HSCT. In subset analysis, we looked at the impact of co-occurring complex cytogenetics on survival outcome of patients in the SH and MH groups. The median OS was 23.6 months, not reached, 20.2 months and not reached in patients with SH and complex cytogenetics, SH without complex cytogenetics, MH with complex cytogenetics and MH without complex cytogenetics, respectively (P=0.18) (Figure 3F). Between patients with SH TP53^MT, those who received intensive chemotherapy induction had significantly better outcome compared to those who received non-intensive chemotherapy with a median OS of 9.97 versus 5.82 months, respectively (P=0.04). However, the benefit of intensive chemotherapy compared to non-intensive chemotherapy in improving OS in MH TP53^MT was less clear with a median OS of 8.03 versus 6.7 months, respectively (P=0.07).

Figure 2.Patterns of the co-mutations identified in the TP53 cohort. Patients with single-hit TP53 are depicted at the Left (black) and patients with multi-hit TP53 at the right (red).

Figure 3.Kaplan-Meier survival curves for patients with single-hit and multi-hit TP53-mutated acute myeloid leukemia. (A) Duration of response, (B) event-free survival and (C) overall survival (OS) in single-hit (SH) versus multi-hit (MH) TP53-mutated acute myeloid leukemia (AML). (D) Subset analysis showing the impact of complex cytogenetics on OS in SH and MH TP53-mutated AML. (E) Landmark analysis for OS among patients with complete remission receiving allogeneic hematopoietic stem cell transplantation in SH versus MH TP53 AML. (F) Landmark analysis for OS among patients with complete remission undergoing allogeneic hematopoietic stem cell transplantation with respect to the presence or absence of complex cytogenetics and TP53 allelic burden. CR: complete remission; CRi: complete remission with incomplete blood count recovery; CK: complex cytogenetics; allo-HCT: allogeneic stem cell transplantation; NR: not reached.

In univariate analysis for OS (Online Supplementary Table S4), age as a continuous variable (every 10 years) (P=0.02), complex cytogenetics (P=0.002), and other low-intensity chemotherapy (P=0.01) were associated with inferior outcomes. RUNX1 mutation (P=0.01), IDH1 mutation (P<0.001), FLT3 ITD mutation (P=0.003), NPM1 mutation (P=0.02), intensive induction (P=0.007) and allogeneic HSCT as a time-dependent co-variate (P<0.001) were associated with favorable OS in univariate analysis. In multivariable analysis for OS, complex cytogenetics (HR=1.56, 95% CI: 1.01-2.40, P=0.04) retained an unfavorable significance, whereas IDH1 mutation (HR=0.24, 95% CI: 0.08-0.71, P=0.01) and allogeneic HSCT (HR=0.28, 95% CI: 0.16-0.47, P<0.001) retained favorable significance.

Discussion

In our real-world, multicenter analysis of a large cohort of patients with TP53^MT AML, we did not observe significant differences in remission rates or survival based on TP53 allelic state. We found that distinct myeloid co-mutation patterns exist between patients with SH and those with MH TP53^MT AML, with IDH1 co-mutations imparting a favorable prognostic significance, and that the use of allogeneic HSCT associating with improved OS, irrespective of SH or MH TP53^MT status.

Recent studies have explored the clinical significance of TP53^MT allelic status in patients with MDS and AML.^4,10,23 While patients with MDS harboring SH TP53^MT tend to have similar outcomes compared to their TP53 wild-type counterparts and better outcomes than those with MH TP53^MT, patients with MDS with excess blasts/AML harboring SH or MH TP53^MT had comparable outcomes. Similarly to Grob et al.,¹⁰ we did not observe significant differences in response rate or survival between patients with SH or MH TP53^MT AML. These data suggest that TP53 allelic state in advanced MDS or AML is less relevant in predicting clinical outcome. Similar to what has been observed in MDS studies, patients with SH TP53^MT AML had an abundance of somatic co-mutations, while MH TP53^MT AML was significantly associated with occurrence of complex cytogenetics.²³

IDH1/2 mutations are observed in approximately 20% of patients with AML (IDH1, 6-16%; IDH2, 8-19%).²⁴ IDH1/2 mutations are more frequently seen in elderly AML patients, especially those with diploid or intermediate-risk cytogenetics, and frequently co-occur with FLT3 ITD and NPM1 mutations.²⁵ With the development of venetoclax and IDH1/2 inhibitors, the outcomes of IDH1/2-mutated AML patients have improved significantly, especially those who are ineligible for intensive therapies.²⁶ Interestingly, we observed significantly improved EFS and OS among patients with IDH1 co-mutations and the favorable significance was retained in multivariate analysis. Moreover, only a small proportion of these patients received venetoclax plus a hypomethylating agent as first-line treatment (2/11 [18%]) or as a salvage therapy (1/11 [9%]) and only 2/11 (18%) patients in this subgroup underwent allogeneic HSCT. None of the patients received an IDH1 inhibitor alone or in combination with chemotherapy upfront. One patient each received a hypomethylating agent/venetoclax plus IDH1 inhibitor and an IDH1 inhibitor alone as a salvage therapy with no response. While these findings are intriguing, they need to be validated in a larger group of patients.

Although allogeneic HSCT is universally considered a potentially curative option for patients with adverse-risk AML, earlier studies showed dismal outcomes for patients with TP53^MT AML undergoing allogeneic HSCT.²⁷ Lack of benefit was attributed to inability to achieve complete response and persistence of the pre-transplant TP53^MT clone. In our earlier report utilizing data from ten US academic centers, we showed that allogeneic HSCT improved survival of patients with TP53^MT AML.¹³ We have now re-confirmed this finding and shown that it holds true irrespective of TP53 allelic state. The multivariable analysis in this study also demonstrated a significantly better EFS associated with induction with a hypomethylating agent plus venetoclax when compared with other therapies. However, this did not translate into improved OS, suggesting evolution of resistant clones that were not suppressed in the long-term by venetoclax plus hypomethylating agent therapy alone, as previously reported.²⁸ Secondly, in a subset analysis we observed better OS with intensive chemotherapy compared to non-intensive chemotherapy induction in the SH and MH subgroups, probably due to the fact that patients eligible for intensive chemotherapy generally have good performance status/fewer co-morbidities and are more likely candidates for allogeneic HSCT. Furthermore, intensive chemotherapy induction did not retain significance for better survival in multivariate analysis.

We acknowledge some limitations of our analysis including a selection bias inherent to a retrospective analysis and some overlap with our prior work.⁹ However, our current analysis includes 382 patients followed longitudinally, significantly strengthening our previous cohort of 291 patients, with more robust TP53 gene annotation data, and these patients have longer follow-up. This strengthened cohort enabled a more comprehensive evaluation of the impact of TP53 mutation burden on clinical outcome. Second, cases with apparent mono-allelic TP53^MT may have hidden clones with bi-allelic TP53 inactivation which were not detected by widely used sequencing methods. Furthermore, although loss of heterozygosity to determine TP53^MT allelic state was not assessed in all patients in this dataset, we defined SH and MH TP53^MT based on earlier observations by Grob et al.¹⁰ Moreover, we did not observe significant differences in survival outcomes using the definition of MH TP53 as per the International Consensus Classification or by Grob et al.¹⁰

In conclusion, unlike in lower-risk MDS, we did not find a significant difference in response rate or survival outcomes between patients with SH or MH TP53^MT AML, which is consistent with recent reports.^10,18 Prospective studies are needed to better understand the effect of TP53 allelic state on the outcomes of patients with TP53^MT AML.

Footnotes

• Received January 6, 2024
• Accepted May 21, 2024

Correspondence

T. Badar badar.talha@mayo.edu

Disclosures

TB has served on advisory boards for Pfizer, Morphosys and Takeda. AP has provided consultancy services for AbbVie; has received research funding from Kronos Bio, Pfizer, Celgene/ BMS, and Servier; has received honoraria from AbbVie and BMS and has received institutional research funding from Pfizer and Kronos Bio. VK has served on advisory boards for Novartis and Pfizer. AMZ is a Leukemia and Lymphoma Society Scholar in Clinical Research; has received institutional research funding from Celgene/BMS, AbbVie, Astex, Pfizer, Medimmune/AstraZeneca, Boehringer-Ingelheim, Cardiff Oncology, Incyte, Takeda, Novartis, Shattuck Labs, Geron, and Aprea; has participated in advisory boards, and/ or had a consultancy role with and received honoraria from AbbVie, Pfizer, Celgene/BMS, Jazz, Incyte, Agios, Servier, Boehringer-Ingelheim, Novartis, Astellas, Daiichi Sankyo, Geron, Taiho, Seattle Genetics, BeyondSpring, Takeda, Ionis, Amgen, Janssen, Genentech, Epizyme, Syndax, Gilead, Kura, Chiesi, ALX Oncology, BioCryst, Notable, Orum, Mendus, Foran, Syros, and Tyme; has served on clinical trial committees for Novartis, AbbVie, Gilead, Syros, BioCryst, Abbvie, ALX Oncology, Geron and Celgene/BMS; and has received travel support for meetings from Pfizer, Novartis, and Cardiff Oncology. The other authors have no conflicts of interest to disclose.