Table 1 Overview on DNMT3A mutation frequencies and prognostic impact in different hematological entities according to the results from this study and previous studies

AML

Ley TJ et al.¹

281 de novo AML pts

62 (22.1%)

WBC count higher in pts with DNMT3A R882 mutation as compared with others (P<0.001)

Association to intermediate KTs; frequency in NK: 44/61; 72.1%; absence of DNMT3Amut in favorable KTs (P=0.001)

Association between DNMT3Amut and FLT3 (P=0.003), NPM1mut (P<0.001) and IDH1mut (P<0.001)

Median OS 12.3 vs 41.1 months in DNMT3Amut vs wt pts (P<0.001); adverse impact also in CN-AML (P=0.007) and in pts with an intermediate risk profile (P=0.006). Adverse impact of DNMT3Amut also in FLT3-ITD-positive pts (P<0.001)

Thol F et al.⁵

489 pts aged 17–60 years: 438 de novo AML; 51 s-AML (excluding acute promyelocytic leukemia)

87 (17.8%)

Higher age (P<0.001), WBC count (P=0.001), and platelet count (P=0.014) in DNMT3Amut than DNMT3A-wt pts

The highest DNMT3Amut frequency was found in pts with CN-AML (71/261; 27.2%)

Pts with DNMT3Amut more frequently had NPM1 (P<0.001), FLT3 (P=0.003) and IDH1mut (P=0.011); more often belonged to the NPM1/FLT3 low-risk group (P<0.001)

Total cohort: DNMT3Amut independently predicted a shorter OS (P=0.005) by multivariate analysis. CN-AML: DNMT3Amut predicted shorter OS (P<0.001) and lower CR rate (P=0.015) as compared with DNMT3A-wt. In more detail, the adverse prognostic impact was only observed in NPM1/FLT3-ITD high-risk group, but not in NPM1/FLT3-ITD low-risk group

Fried I et al.⁹

98 s-AML/t-AML (37 s-AML, 61 t-AML)

Not analyzed

Higher frequency of DNMT3Amut in CN-AML (11/25; 44%) as compared to cytogenetically aberrant pts (7/50; 14%; P=0.008)

IDH1/2mut associated with DNMT3Amut (P=0.007)

DNMT3Amut pts had a shorter interval between MDS/MPN diagnosis and transformation to s-AML (median time to progression, 7.5 vs 12 months; P=n.s.)

Marcucci G et al.¹⁰

415 CN-AML

142/415 (34.2%)

Younger (<60 years; n=181) and older (⩾60 years; n=234) pts had similar frequencies of DNMT3Amut (35.3 vs 33.3%). DNMT3Amut pts had higher WBC counts (P<0.001) and BM blasts (P=0.03)

(Only CN-AML patients investigated)

DNMT3Amut pts harbored NPM1mut (P<0.001) and FLT3-ITD (P=0.01) more often and CEBPAmut (P<0.001) less often than DNMT3A-wt

Total cohort: DNMT3A-mut pts had shorter DFS (P=0.03) and OS (P=0.07) than wt pts. In multivariate analyses, DNMT3Amut remained associated with shorter DFS (P=0.01). Age-specific associations: Younger pts with non–R882-DNMT3Amut had shorter DFS (P=0.002) and OS (P=0.02), whereas older DNMT3Amut pts had shorter DFS (P=0.005) and OS (P=0.002) after adjustment for other prognosticators

Renneville A et al.¹¹

123 adults aged 16–60 years with de novo CN-AML

36/123 (29.3%)

DNMT3Amut and wt cases did not differ significantly by age, gender and WBC count. 55% of DNMT3Amut cases were FAB M4 or M5 vs 21% of DNMT3A-wt (P<0.001)

(Only CN-AML patients investigated)

DNMT3Amut were associated with NPM1mut (P=0.0006). None of the 36 (0%) DNMT3Amut pts had a concomitant CEBPAmut (P=0.017)

DNMT3Amut pts had shorter 5-year EFS (13 vs 32%; P=0.02) and -OS (23 vs 45%; P=0.02) than DNMT3A-wt pts. In multivariate analysis, DNMT3Amut remained an independent adverse prognostic factor for EFS and OS

This study

194 CN-AML

70/194 (36.1%)

Significant association with female gender (P<0.001) and younger age (P=0.021)

(Only CN-AML pts investigated)

DNMT3Amut were associated with NPM1mut (P<0.001), FLT3-ITD (P=0.08) and IDH1mut (P=0.002). DNMT3Amut less frequent with ASXL1 (P=0.001) and RUNX1 (P=0.002)

No significant impact of DNMT3Amut on survival outcome (mutated vs wt: 2-year OS: 60.0 vs 58.0%; median OS: 62.2 months vs 35.6 months; P=n.s.; 2-year EFS: 34.2 vs 42.1%; median OS: 13.6 months vs 14.9 months; P=n.s.)

MDS

Walter MJ et al.¹²

150 de novo MDS

12/150 (8.0%)

No significant association between DNMT3Amut and gender, age, bone marrow blasts, blood counts or MDS subtypes

Not analyzed

Not analyzed

DNMT3Amut pts had worse OS than pts without (P=0.005) and more rapid progression to s-AML (P=0.007). 7/12 (58.3%) pts with DNMT3Amut progressed to s-AML compared to 39/138 (28.3%) pts without (P=0.007)

Thol F et al.¹³

193 MDS

5/193 (2.6%); no correlation with MDS subtype

Not analyzed

NK in three of five DNMT3Amut pts

One DNMT3Amut pt had a concomitant ASXL1mut, and one had a mutation in IDH1

3/4 DNMT3Amut pts (75.0%) with follow-up data developed s-AML compared to 28.2% pts with DNMT3A-wt (P=0.043)

This study

115 de novo MDS

15/115 (13.0%); no correlation with MDS subtype

No associations

NK in 9/15 DNMT3A-mutated pts

No significant associations of DNMT3Amut with any specific mutation

DNMT3Amut had no significant impact on OS and EFS

CMML

Jankowska AM et al.¹⁴

52 CMML; 20 s-AML following CMML

CMML: 2/52 (3.8%); s-AML: 6/20 (25.0%)

Total cohort: increased marrow blasts (P=0.019) and higher WBC counts (P=0.016) in DNMT3Amut pts

Most DNMT3Amut cases had a NK (5/7 with available cytogenetics; 71%)

DNMT3Amut were the sole abnormality in three pts; in four cases associated with other mutational events

Total cohort: blast count, disease stage and DNMT3Amut status (P=0.04) had a significant effect on survival. By multivariate analyses only disease stage (CMML-1/2 vs s-AML) remained relevant

This study

103 cases (65 CMML-1, 38 CMML-2)

Total cohort: 7 (6.8%); CMML-1: 2 (3.1%); CMML-2: 5 (13.2%)

Significant association of DNMT3Amut with female gender (P=0.029)

NK in 5/6 DNMT3Amut pts

4/6 DNMT3Amut cases showed concomitant mutations (no association with specific mutations)

DNMT3Amut showed neither impact in the total cohort (P=n.s.), nor in the cohort of CMML-2 cases (P=n.s.)

T-ALL

This study

99 T-ALL

17/99 (17.2%)

Significant association of DNMT3Amut with higher age (P<0.001) and lower hemoglobin level (P=0.02)

Association of DNMT3Amut with NK (P=0.035)

No association of DNMT3Amut with distinct mutation

DNMT3Amut had an inferior OS compared to wt pts (2-year OS: 35.2 vs 67.9%; median OS: 16.6 months vs n.r.; P=0.013)