Characteristics and Prognosis of "Acute Promyelocytic Leukemia-like" Nucleophosmin-1-Mutated Acute Myeloid Leukemia in a Retrospective Patient Cohort

Abstract

Background: AML with NPM1 mutation is the largest subcategory of AML, representing about 35% of AML cases. It is characterized by CD34 negativity, which suggests a relatively differentiated state of the bulk of leukemic blasts. Notably, a significant subset of NPM1-mutated AML cases also exhibit HLA-DR negativity, classifying them as "double-negative", and mimicking, therefore, the CD34^- HLA-DR^- immunophenotype of acute promyelocytic leukemia (APL). Objectives: This study focuses on the "acute promyelocytic leukemia-like" ("APL-like") subset of NPM1-mutated AML, which can be challenging to distinguish from APL at presentation, prior to confirming RARa translocations. We aim to investigate the hematologic and immunophenotypic parameters that may aid to its distinction from APL. Additionally, we explore differences in genetic profile and prognosis between "APL-like" and "non-APL-like" NPM1-mutated AML cases. Methods: We conducted a retrospective evaluation of 77 NPM1-mutated AML cases and 28 APL cases. Results: Morphological characteristics, hematologic parameters (such as DD/WBC and PT/WBC), and specific immunophenotypic markers (including SSC, CD64, and CD4) can assist in the early distinction of "APL-like" NPM1-mutated AML from APL. Regarding differences in genetic profiles and outcomes between "APL-like" and non-"APL-like" NPM1-mutated AML cases, we observed a significantly higher incidence of IDH1/2 /TET2 mutations, along with a significantly lower incidence of DNMT3A mutations in the "APL-like" subset compared to the non-"APL-like" subset. The frequency of Ras-pathway and FLT3 mutations did not differ between these last two groups, nor did their prognoses. Conclusions: Our findings contribute to a comprehensive characterization of NPM1-mutated AML, enhancing diagnostic accuracy and aiding in the detailed classification of the disease. This information may potentially guide targeted therapies or differentiation-based treatment strategies.

Keywords: AML with NPM1 mutation; DNMT3A mutations; FLT3 mutations; IDH1/2 mutations; Ras-pathway mutations; TET2 mutations; acute promyelocytic leukemia.

Figure 1

Morphology of peripheral blood (PB) blasts in APL and in “APL-like” NPM1m AML. (A) Peripheral blast of patient with APL, featuring a typical bilobed nucleus with typically overlapping lobes and an inconspicuous Auer rod, as is often the case in APL in PB. (B) Peripheral blasts in “APL-like” (CD34- HLADR-) NPM1-mutated AML case, displaying a prominent Auer rod and a tendency to “cup-like” nuclei. (C) Peripheral blast from another patient with APL, showing a typical bilobed nucleus with typically overlapping lobes and non-readily visible Auer rods. (D) Peripheral blasts in an “APL-like” (CD34- HLADR-) NPM1-mutated AML case, characterized by typical “cup-like” nuclei.

Figure 2

Distribution and differences in DDs/WBC and PT/WBC ratios between APL and “APL-like” NPM1m AML (***: p < 0.001). (A) Box plot of ratios DDs/WBC (mg/L / G/L) in our acute promyelocytic leukemia (APL) cases and “APL-like” NPM1-mutated cases (APL-like NPM1m). Difference is statistically significant between the two groups (p < 0.001, Wilcoxon’s rank-sum test). (B) Box plot of ratios PT/WBC (s/G/L) in our acute promyelocytic leukemia (APL) cases and “APL-like” NPM1-mutated cases. Difference is statistically significant between the two groups (p < 0.001, Wilcoxon’s rank-sum test). (C) DDs/WBC ratios of APL cases shown with black circles, and DDs/WBC ratios of “APL-like” NPM1-mutated cases are shown with red triangles. Cases with ratio > 4.92 (mg/L/G/L) are always APL. (D) PT/WBC ratios of APL cases are shown with black circles, and PT/WBC ratios of “APL-like” NPM1-mutated cases are shown with red triangles. Cases with ratio > 8.54 (s/G/L) are always APL.

Figure 3

Differences in expression of immunophenotypic markers between APL and “APL-like” NPM1m-AML. (A–C) The expression levels of CD117, MPO, and CD38 did not differ significantly between acute promyelocytic leukemia (APL) cases and “acute promyelocytic leukemia-like” NPM1-mutated AML cases (ns: non-significant, **: p < 0.01, ***: p < 0.001). (D–F) Significant differences were observed in the expression of CD4 and CD64 and in SSC values between the two cohorts, with APL cases showing higher CD64 expression and SSC values, and “APL-like” NPM1m AML cases showing higher CD4 expression. Comparisons were made using Wilcoxon’s rank-sum test. It is important to note that “APL-like” NPM1-mutated AML, as immunophenotypically defined in this study (CD34- HLADR-), is “by definition” CD64-negative (CD64-positive NPM1-mutated cases exhibit a monocytic phenotype expressing HLA-DR).

Figure 4

Differences in the proportion of mutated versus wild-type cases for genes IDH1/2/TET2, “Ras-pathway genes”, DNMT3A, and FLT3 among “APL-like” and “non-APL-like” NPM1m AML. ns: non-significant, *: p < 0.05, **: p < 0.01. (A) Proportion of IDH1/2/TET2-mutated cases among “APL-like” NPM1m AML and “non-APL-like” NPM1m AML. “APL-like” NPM1m AML exhibits a significantly higher rate of IDH1/2/TET2 mutations (chi-square, p = 0.0143). (B) Proportion of Ras-pathway-mutated cases among “APL-like” NPM1m AML and “non-APL-like” NPM1m AML. No significant difference in the frequency of Ras-pathway mutations (chi-square, p = 0.1391) was observed between the two entities. (C) Proportion of DNMT3A-mutated cases among “APL-like” NPM1m AML and “non-APL-like” NPM1m AML. “APL-like” NPM1m-AML shows a significantly lower rate of DNMT3A mutations (chi-square, p = 0.0018). (D) Proportion of FLT3-mutated cases among “APL-like” NPM1m AML and “non-APL-like” NPM1m AML. No significant difference in the frequency of FLT3 mutations (chi-square, p = 0.4344) was found between the two entities.

Figure 5

Event-free survival (EFS) and overall survival (OS) of “APL-like” and “non-APL-like” NPM1m AML in our cohorts. (A) Event-free survival: no statistically significant difference was observed between “APL-like” and “non-APL-like” NPM1-mutated AML cases (p = 0.5 log-rank test). (B) Overall survival: no statistically significant difference was observed between “APL-like” and “non-APL-like” NPM1-mutated AML cases (p = 0.7, log-rank test).