The clinicopathologic significance of NPM1 mutation and ability to detect mutated NPM1 by immunohistochemistry in non-AML myeloid neoplasms

Abstract

NPM1 mutated non-AML myeloid neoplasms (MN; <20% blasts) are characterized by an aggressive clinical course in a few studies. In this retrospective study, we evaluate the clinicopathologic and immunohistochemical features of non-AML MN patients with NPM1 mutations. We assessed NPM1 mutation by targeted next generation sequencing (NGS). Cytoplasmic NPM1 expression was assessed by immunohistochemistry (IHC) on formalin-fixed, formic acid-decalcified bone marrow biopsy specimens. We evaluated 34 non-AML MN patients with NPM1 mutations comprising MDS (22), MPN (3) and MDS/MPN (9). They commonly presented with anemia (88%), thrombocytopenia (58%) and leukopenia (50%). Bone marrow dysplasia was common (79%). The karyotype was often normal (64%). NGS for MN-associated mutations performed in a subset of the patients showed a median of 3 mutations. NPM1 mutations were more often missense (c.859C > T p. L287F; 65%) than frameshift insertion/duplication (35%) with median variant allele frequency (VAF; 9.7%, range 5.1%–49.8%). Mutated NPM1 by IHC showed cytoplasmic positivity in 48% and positivity was associated with higher VAF. The median overall survival (OS) in this cohort was 70 months. Nine patients (26%) progressed to AML. OS in patients who progressed to AML was significantly shorter than the one of patients without progression to AML (OS 20 vs. 128 months, respectively, log rank p = 0.05). NPM1 mutated non-AML MN patients commonly had cytopenias, dysplasia, normal karyotype, mutations in multiple genes, and an unfavorable clinical outcome, including progression to AML. Our data demonstrated that IHC for NPM1 can be a useful supplementary tool to predict NPM1 mutation in some non-AML MN; however, genetic testing cannot be replaced by IHC assessment.

1 |. INTRODUCTION

The NPM1 gene encodes a multifunctional nucleocytoplasmic shuttling protein located on chromosome 5q35 and contains 12 exons (Figure 1).¹ NPM1 protein shuttles between the nucleus and cytoplasm, acting as a molecular chaperone that facilitate multiple protein–protein interactions.² Functions of NPM1 include ribosome biogenesis, protein synthesis, cell growth, cell proliferation, nuclear export of the ribosomal protein L5/5S rRNA subunit complex, maintenance of genomic instability, mediation of apoptosis and control of centrosome duplication.^1–5

FIGURE 1.

Genetic characteristics of non-AML myeloid neoplasm (MN) patients with NPM1 mutations. (A) Structure and functional domains of the NPM1 gene. At the N-terminus, the protein displays two nuclear export signal (NES) motifs (residues 42–49 and 94–102); centrally, there is a metal binding domain, two acidic regions (residues 120–132 and 160–188), a bipartite nuclear localization signal (NLS) motif (residues 152–157 and 190–197), a basic cluster inside a moderately basic region. At the C-terminus an aromatic region unique to NPM isoform 1 containing the nucleolar localization signal (NLS) with tryptophan residues 288 and 290 is present. (B) Summary table of genetic characteristics of non-AML MN patients with NPM1 mutations. (C) Multiple and different genetic mutations were observed in 11 patients who had a myeloid gene sequencing done; the median number of genes seen in each patient was 3. (D) Dot plot graph showing that patients with NPM1 missense mutations had a significantly lower variant allele frequency (%) compared to frameshift genetic alterations.

NPM1 mutations have been identified in a number of hematological neoplasms including various myeloid neoplasms (MN).^1,6NPM1 mutations are founder lesions and one of the most common somatic mutations (approximately 30%) identified in de novo AML; up to 50% of AML with a normal karyotype (AML-NK) harbor an NPM1 mutation.^7,8NPM1 mutations have been found to be more common in de novo AML compared to secondary AML.⁹NPM1 mutations in AML-NK, in the absence of FLT3 mutations, usually show favorable prognosis among this heterogeneous disease category.^1,10 Interestingly, NPM1 mutations have been identified in a small fraction of patients with secondary AML, MDS (2%), and MDS/MPN (3%).^7,11 In MDS, most patients with NPM1 mutations were subcategorized as MDS with excess blasts (EB); in MDS/MPN patients, a large fraction of patients were diagnosed as chronic myelomonocytic leukemia (CMML).^8,11 23% of patients with non-AML MN with NPM1 mutations were found to present with an abnormal karyotype; further, NPM1 mutation was found to be not the sole genetic mutation present.¹² Overall, the genetic factors that lead to evolution of non-AML MN to AML remain poorly understood and more studies are necessary to understand the factors that lead to disease progression in these patients.¹³

NPM1 immunohistochemistry (IHC) was first investigated in AML as a possible surrogate marker and was found to localize aberrantly in the cytoplasm; while wild type NPM1 was predominantly located in nuclei and nucleoli.¹⁴ The immunostaining was shown to be highly specific to AML when compared to other hematological neoplasms.¹⁵ Further, the NPM1 immunostaining also highly correlated with NPM1 mutations in AML, thus adding this marker as a possible diagnostic tool.^9,14,15 It is important to note that few studies have demonstrated that NPM1 IHC does not correlate with the NPM1 mutational status,¹⁶ raising concerns about the diagnostic utility of NPM1 IHC in AML patients. In non-AML MN (<20% blasts) such as MDS, some studies have suggested that NPM1 staining also highly correlates with NPM1 mutations.⁹ However, because NPM1 mutation positive non-AML MN are rare, more studies are essential to further assess the correlation between NPM1 mutational status and NPM1 IHC.

NPM1 mutations in non-AML MN have been investigated in only some studies;^7,8 with the mutation being associated with poor prognosis.⁷ This observation is interesting because NPM1 mutations in de novo AML are associated with a relatively favorable prognosis. In this study we identified non-AML MN patients with NPM1 genetic mutations. We utilized a mutated (m) NPM1 antibody that does not cross-react with wild-type NPM1 or unrelated cellular proteins.¹⁷ Our main objective was to assess the relationships between NPM1 mutations and NPM1 IHC in this patient cohort. We also studied the role of NPM1 mutations on disease course and prognosis of these patients.

2 |. MATERIALS AND METHODS

2.1 |. Patients

After Institutional Review Board approval and in accordance to ethics committee approvals of the Cleveland Clinic Foundation, archived bone marrow (BM) biopsy reports and molecular genetic results between 2010 and 2019 were searched for positive NPM1 mutation patients. Then, clinical and pathologic data in patients with NPM1 mutation and non-AML MN, defined by the presence of less than 20% peripheral blood (PB) or BM blasts were reviewed. Patients’ characteristics are shown in Table 1. All cases were diagnosed according to the 2016 WHO classification. In cases with available paraffin embedded core biopsy and/or clot section blocks, immunohistochemistry for mutated NPM1 was performed.

TABLE 1.

Clinicopathological characteristics of non-AML myeloid neoplasm patients with NPM1 mutations (n = 34).

Patient characteristics
Age	65.5 years (median)
Gender	Male: 22/Female: 12
Laboratory parameters
Anemia	88%
Hb	9.2 g/dL (median)
Thrombocytopenia	58%
Platelet	98.5 × 103μL (median)
Leukopenia	50%
WBC	3.85 × 103μL (median)
Diagnosis	MDS: 22 patients
	MPN: 3 patient
	MDS/MPN: 9 patients (including4 CMML)
Bone marrow cellularitya	Hypercellular: 21 Hypocellular: 4 Normocellular: 6
Bone marrow blasts	3.5% (median; range 0–18)
Bone marrow dysplasia	Overall: 79%
	• Megakaryocyte: 60% • Erythroid: 50% • Granulocyte: 35% Multilineage dysplasia 44%
Bone marrow fibrosis	26%

^aBone marrow aspirate/ biopsy specimens were either not available (1 patient) or suboptimal (2 patients).

2.2 |. Targeted next generation sequencing studies

Results of various targeted sequencing platforms (TruSight, TruSeq and Nextera) (Illumina) to detect somatic variants were evaluated. Sequencing libraries were generated according to an Illumina paired-end library protocol. The enriched targets were sequenced using a HiSeq 2000 or MiSeq (Illumina). Variants were annotated using Annovar. Variants with minimum depth less than 20 or number of high-quality reads less than five were filtered out. Average coverage was at least 250x. A bio-analytic pipeline developed in-house identified somatic/germline mutations using sequences derived from controls and mutational databases such as dbSNP138, 1000 Genomes or ESP 6500 database, and Exome Aggregation Consortium (ExAC) to exclude germline variants.^18–20

2.3 |. Immunohistochemistry

Expression of mutant NPM1 protein using a mutation specific antibody mNPM1 was done. Positive and negative control samples were assessed to validate the mNPM1 antibody staining patterns. Formalin-fixed, paraffin-embedded tissue sections of BM biopsy specimens were deparaffinized in xylene and then rehydrated through graded concentrations of alcohol. The staining was performed on a Ventana Discovery Ultra and staining visualized using the ChromoMap Detection Kit (Ventana Medical Systems #760–159). Blocking was performed with Agilient Technologies Serum Free Protein Block (#X090930–2) for 16 minutes, followed by epitope retrieval for 24 minutes (Ventana Medical Systems Ultra Cell Conditioning CC2 #950–223). The mNPM1 antibody (Thermo Fisher Scientific; PA146356) was diluted 1:2000 and incubated for 16 minutes. Then secondary antibody incubation was done with OmniMap AntiRabbit (Ventana Medical System #760–4311) for a 12 minutes incubation. Finally, hematoxylin and Bluing solutions incubations were done for 4 minutes each. Slides were then air dried and cover-slipped. IHC Positivity was determined when an aberrant nuclear and cytoplasmic localization is found.

2.4 |. Statistical analyses

Statistical analysis was performed using MedCalc software (version 19.1.5) and Microsoft Excel 2013 software. Categorical variables were summarized using frequencies and percentages, while continuous variables were summarized using medians and ranges. For categorical variables, Fisher’s exact and Mann–Whitney U test were used to compare groups. For continuous variables, independent t-test was performed to compare groups. All p values were two-tailed, and p < 0.05 was considered significant. The Kaplan–Meier method was used for overall survival (OS) estimates (Log rank test) from the date of initial diagnosis to the most recent follow-up or the time of death.

3 |. RESULTS

3.1 |. Patient demographics

We identified 34 patients (Male: 22/Female: 12) with non-AML MN and NPM1 mutations. The overall clinicopathological characteristics of our patient cohort is shown in Table 1. The majority of the patients presented with anemia (88%; median 9.2 g/dL), leukopenia (50%; median 3.85 × 10³/μL), and/or thrombocytopenia (58%; median 98.5 × 10³/μL). Initial BM assessment in our cohort showed that 65% of patients presented with hypercellular marrow for age. The median BM blast percentage based on morphology was 3.5% (range 0–18%). The majority of the patients presented with BM dyspoiesis (79%), with 44% of the patients showing multilineage dysplasia (Table 1). Some of the patients had BM fibrosis. Based on the BM morphology, cytogenetics, and molecular studies, the patient’s diagnoses were as follows: MDS (22/34), MPN (3/34), and MDS/MPN (9/34) (Table 1).

3.2 |. Cytogenetic and molecular studies

Most patients had a normal karyotype (21/33; 64%). Among those with abnormalities (12/33; 36%), seven had a single abnormality, two had two abnormalities and only one had a complex karyotype. Common abnormalities were del(20q) (4/12) and del(5q) (2/12) (Figure 1). MDS fluorescence in situ hybridization (FISH), including probes for 5q, 7q, 20q and centromeric chromosome 8, was performed at our institution for subset of patients; 4/13 patients had an abnormal FISH result; two cases had deletion 20q, one had deletion 5q, and one had monosomy 5. Six patients had a myeloid whole exome sequencing panel (62 genes) performed at our institution. The overall number of genes mutated were 11, with each patient presenting with a median of three genetic mutations (Figure 1). NPM1 mutational analysis showed that the majority of patients had a somatic c.859C > T mutation (p. L287F; 65%), and a subset of patients showed frameshift insertion (ins) or other duplications (dup) or ins. The overall mean variant allele frequency (VAF) was 9.7% (range 5.1%–49.8%). VAF% in ins or dup was significantly higher than c.859C > T mutation (mean 26.2% vs. 8.1%, p = 0.0002) (Figure 1).

3.3 |. mNPM1 Immunohistochemistry

A total of 31 patients with NPM1 mutations had mNPM1 IHC performed (Table 2) with appropriate controls. Representative cases with positive or negative mNPM1 is shown in Figure 2. A total of 48% of the patients with NPM1 mutations (n = 15) were positive for mNPM1 by IHC. There was no predilection for MDS versus non-MDS disease status. Patients with positive mNPM1 immunostain had higher blast count at presentation than patients with negative mNPM1 (Table 2). Cells expressing mNPM1 included myeloid blasts, immature myeloid cells, pronormoblasts, and megakaryocytes (subset). Patients with positive mNPM1 immunostain had c.859C > T mutation (3 cases) and other ins or dup (12 cases). Further, we found that patients negative for mNPM1 by immunostain all showed c.859C > T mutation (16 cases) (p < 0.001).

TABLE 2.

Clinicopathological findings of non-AML myeloid neoplasm patients with NPM1 mutations based on mNPM1 Immunohistochemistry.

	NPM1 IHC positive	NPM1 IHC negative	p value
Diagnosis	15	16
	• 10 MDS	• 9 MDS
	• 5 non-MDS	• 7 non-MDS
BM blasts	7% (median)	2% (median)	0.01a
NPM1 mutation	c.859C > T mutation: 3 cases	c.859C > T mutation: 16 cases	<0.001b
	Insertion or duplication: 12 cases
VAF%	18.9% (median)	8.06% (median)	0.01a
Progress to AML	5/15 (33%)	4/16 (25%)
Abnormal karyotype	6/15 (40%)	5/15 (33%)
Survival	20 months (median)	48 months (median)

^at-test.

^bChi-square test.

FIGURE 2.

Representative positive or negative mNPM1 immunohistochemistry in four non-AML myeloid neoplasm patients. (A–D) Upper panels show hematoxylin and eosin stain and lower panels show mNPM1 immunohistochemistry. (A) Patient showing negative mNPM1 with a diagnosis of chronic myelomonocytic leukemia (CMML)-1 with bone marrow (BM) blasts 8%, NPM1 mutation c.859C > T (p.L287F; VAF 7.02%) and normal male karyotype (46,XY[20]). (B) Patient showing positive mNPM1 with a diagnosis of myelodysplastic syndrome with excess blasts (BM blasts 7%), NPM1 mutation c.859C > T (p.L287F; VAF 7.94%) and abnormal karyotype (46,XY,add(4)(p11)[2]/46,XY[19]). (C) Patient showing positive mNPM1 with a diagnosis of CMML-2 with BM blasts 17%, NPM1 genetic duplication c.860_863dupTCTG (p.W288Cfs; VAF 29.90%) and normal male karyotype (46,XY[20]). (D) Patient showing positive mNPM1 with a diagnosis of myelodysplastic syndrome with BM blasts 3%, NPM1 duplication c.860_863dupTCTG (p.W288Cfs; VAF 31.20%) and abnormal karyotype (46,XX,del(5)(q15q33)[4]/46,XX[16]).

Patients with positive mNPM1 IHC had a higher VAF % than patients with negative mNPM1 IHC (mean 20.8% vs. 8.4%, respectively; p = 0.0015). Among positive mNPM1 IHC cases, the majority of the cases had a VAF of 10% or higher (12/15, 80%). Patients with negative mNPM1 IHC had a VAF less than 10% in the majority of the cases (13/16, 81%). A VAF cut off of 10% shows a positive predictive value of 80% and a sensitivity of 80%. Lastly, we did not observe any statistical difference between mNPM1 positive and negative patients as regards disease progression, abnormal karyotype and OS (Table 2).

3.4 |. Management and clinical outcome

Follow-up of the non-AML MN patients with NPM1 mutations showed that most of these patients received supportive treatment (29%) and/or therapy (medical treatment or BM transplantation (BMT; 65%); a small proportion were lost to follow-up (6%). Supportive treatment received by patients was mainly red blood cell (RBC) and/or platelet transfusions. Treatment strategies included hypomethylating agents (55%), chemotherapy (7%), and/or bone marrow transplantation (24%). The median OS in this cohort was 70 months. There is no difference in OS between patients with MDS and non-MDS including MPN and MDS/MPN (70 vs. 117 months, respectively, log rank p = 0.82).

Nine patients in our cohort had disease progression to AML; one patient had increasing blasts not meeting criteria for AML (<20%). Of the nine patients that transformed to AML, six patients had MDS-EB, one patient had MDS-multilineage dysplasia, one patient had MDS/MPN-unclassifiable and one patient had CMML-2 (Table 3). The majority of the patients that progressed to AML had a disease relapse (7/9; 78%). The OS for patients that had AML disease progression was significantly shorter compared to patients without disease progression (AML transformed OS: 20 months; non-AML OS: 128 months; log rank p = 0.05).

TABLE 3.

Management and clinical outcome of non-AML myeloid neoplasm patients with NPM1 mutations.

Management, follow-up and survival
Treatment	Supportive treatment: 29% (10/34) Other therapy: 65% (22/34) • Hypomethylating agent: 55% • Chemotherapy: 7% • BM transplant: 24% Lost to follow-up: 6% (2/34)
Transformation to AML or increase in blasts	9 patients transformed to AML • 6MDS-EB • 2 MDS-EB1 4 MDS-EB2 • 1CMML-2 • 1 MDS-MLD • 1 MDS/MPN-U 1 patient had an increase in blasts (<20%) • MDS(RCMD)
AML relapsea	Positive: 78% (7/9) Negative: 11% (1/9) Not known: 11% (1/9)
Follow-up	29.5 months (median); 41.7 months (mean)
Survival	Total median OS(n = 34): 70 months OS in patients with progression to AML (n = 9): 20 monthsb OS in patients without progression to AML (n = 25): 128 monthsb

^aPatients who progressed to AML and received appropriate treatment had disease relapse after initial response. Most patients received chemotherapy treatment.

^blog rank p = 0.05.

4 |. DISCUSSION

In this study, we assessed the clinicopathologic and genetic features of a cohort of patients of non-AML MN with NPM1 mutations. MDS comprised the vast majority of the patients in our cohort (22 patients), followed by MDS/MPN (nine patients; including four CMML) and lastly MPN (three patients). These results are overall representative of the case distribution that we see in our clinical practice; MDS being the most common non-AML MN observed. While research has previously demonstrated that NPM1 mutations confer intermediate to good prognosis in AML; NPM1 mutations are likely an adverse factor in the non-AML NPM1-mutated MNs.⁷ In our study, non-AML MN patients with NPM1 mutations commonly presented with cytopenias, dysplasia, normal karyotype, mutations in multiple genes, and an unfavorable clinical outcome, including a high rate of progression to AML. Further, our results have shown that the majority of the patients who progressed to AML had disease relapse/progression (78%; 7/9). This observation seems to concur with Patel et al.⁷ observations which demonstrated that patients with non-AML NPM1 mutations progressed more quickly to AML than other patients. Forghieri et al.⁹ also observed that non-AML MN patients with NPM1 mutations also have a higher rate of disease progression to AML and further suggested that MNs with NPM1 mutations may best be classified as AML even with <20% blasts. We partially agree with these observations; however, we believe the genetic heterogeneity of MNs and the numerous confounding factors affecting disease progression necessitate that further studies be done to assess the role of NPM1 in non-AML MN. In MDS patients with NPM1 mutations, we observed a significant number of patients categorized as MDS-EB (6/9); our results were different than the results of Dicker et al.²¹ who observed none. Further, we observed NPM1 mutations in a few cases with associated distinct MDS cytogenetic alterations, an observation that Dicker et al.²¹ did not observe, suggesting that NPM1 mutations do not solely present in MDS patients with normal karyotype. The differences in disease presentation can be explained by the disease heterogeneity of MN. Bejar et al.²² assessed the role of different genes in MDS, the most common non-AML MN and suggested that ASXL1, RUNX1, TP53, EZH2, CBL, and ETV6 genetic alterations might be associated with adverse prognosis in this disease category. We and others believe that the constellation of the different genetic alterations together can aid in suggesting prognosis in non-AML MN.²² Our results demonstrated that the majority of non-AML patients with NPM1 mutations presented with a normal karyotype, with only few cases demonstrating significant MDS cytogenetic alterations. Different genetic mutations were observed in the cases of non-AML NPM1-mutations, including: DNMT3A, WT1, TET2, KRAS, PTPN11, BCOR, IDH2, RAD21, RIT1, and TP53. Most cases presented with three genetic mutations. In comparison to NPM1 mutant AML where FLT3 mutations do occur; our cohort of patients did not show FLT3 mutations, suggesting that this finding is extremely rare in non-AML MN patients with NPM1 mutations. FLT3 genetic alteration might be a late event in disease progression predominantly affecting AML patients only.²³ In addition, we noted that DNMT3A mutations were the most associated lesions with NPM1 mutation in non-AML MNs; however, due to the paucity of the cases, further studies are needed to confirm this association. In our study, we observed the presence of c.859C > T mutation which is uncommon in AML; in AML the most common mutation being TCTG frameshift insertions (c.860–863dup p.W288Cfs). Therefore, the c.859C > T missense mutation in non-AML MN could be a passenger mutation rather than a founder genetic mutation especially taking into consideration that the mutation has been associated with low VAFs (Figure 1). Larger studies comparing the NPM1 mutations in AML and non-AML MN is warranted to further understand the role of the different NPM1 mutations in MN; further, functional studies will aid in developing a more comprehensive understanding of the role of this interesting finding. Overall, we believe that NPM1 c.859C > T mutation is likely pathogenic as reported across some genetic databases; however, the small sample size and the lack of functional studies precludes a definitive conclusion regarding this finding.

Immunohistochemistry is a relatively inexpensive and readily available diagnostic tool compared to molecular studies. As previously discussed in cases of AML with NPM1 mutations, there was a high correlation between the mutational status and positive cytoplasmic staining^1,15 with only rare studies showing a poor correlation.¹⁶ In the current study, only a subset of non-AML MN patients with NPM1 mutations showed mNPM1 positivity by IHC for mutant protein (~50%). Our results appear similar to the conclusions suggested by Konoplev et al.¹⁶ that suggested that NPM1 mutations do not always correlate with NPM1 IHC in AML. The mNPM1 IHC positivity shown in a subset of cases with NPM1 mutations might be explained by the differences in genetic mutations, VAF or other upstream or downstream events affecting gene expression. It is likely that the missense mutation c.859C > T might be a contributing factor to the discrepancy that we observed between the genetic findings and mNPM1 IHC. We did not find differences in mNPM1 protein expression based on disease status (AML vs. non-AML), disease progression, abnormal karyotype and survival. Forghieri et al.⁹ and others have suggested that MDS patients with positive mNPM1 have a worse prognosis (OS and disease progression to AML) than patients with negative staining. In our study, we observed a difference in the survival and disease progression between mNPM1 positive and negative groups; however, the difference was not statistically significant possibly due to the small sample size (Table 2). Interestingly, we found that all the mNPM1 negative cases harbored c.859C > T mutation (16/31; 52%; p < 0.001) suggesting that this mutation might be leading to a non-expressed protein compared to the other genetic mutations. Another possible explanation for the differences in mNPM1 IHC might be due to the differences in the VAF of the NPM1 mutations. We observed a statistically significant difference in VAF between mNPM1 IHC positive and negative cases (18.9% vs. 8.06%; p = 0.01), suggesting that high mutational burden is associated with positive mNPM1 IHC. Patel et al.²⁴ has observed a correlation between high VAF and decreased OS in patients with de novo AML, suggesting that high tumor burden of NPM1 mutations could be of diagnostic significance. Further, differences in VAF in some MN have been associated with differences in clinical outcome and complications.²⁵ Although, drawing conclusions regarding the correlation between IHC and VAF is informative, there are some technical caveats such as discordance between gene mutation and protein expression. In other words, NPM1 mRNA expression levels might not be reflected in the level of protein detected on IHC. This might be due to many causes. Firstly, the variations in the mutant cell population between the slide and the aspiration sample; secondly the gene expression of the mutant population might be affected by the expression of non-mutant population and the background of stromal cells and inflammatory cells; lastly, the library preparation for NGS assay can be a confounding factor in generating a VAF. To be able to do a more robust correlation, a standardized quantitative method for mNPM1 IHC should be devised as previously shown by Patel et al.^7,24 Overall, we believe that the VAF in non-AML MN patients with NPM1 mutations is a useful information in these patients that can help predict the disease progression in some patients.

The best approach to study disease progression in non-AML MN would be to serially evaluate the genetic alterations in the different stages of disease evolution.¹³ In this study we assessed in depth the effects of the different NPM1 mutations and the utility of IHC in assessing disease progression, similar to the study design as Forghieri et al. and others.^7,9 That the understanding of the association between mNPM1 IHC patterns and localization in NPM1 non-AML MN patients has been lacking⁹ prompted us to design the current study. Our results suggest that assessing NPM1 mutations especially in non-AML MN with normal karyotype is very useful in excluding the mutation in these patients,⁹ and that IHC alone is not sufficient to rule out the mutation because different mutations and tumor burden seem to effect the mNPM1 IHC expression. Some limitations of our current study include the sample size, the retrospective study design and the collection of the specimens which is derived from a single institution. Further, functional studies will be necessary to understand how the different NPM1 gene mutations affect disease pathogenesis and protein expression. Our results provide further evidence that NPM1 mutations in non-AML MN patients is a possible significant adverse prognostic factor that might necessitate a more aggressive therapeutic approach especially in the context of other risk factors such as age and overall performance status.⁷ However, the complex heterogeneous mutational profiles of the MDS phenotype in specific and secondary AML make studying the effect of individual gene mutations difficult.¹³

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.