NPM1 mutations are more stable than FLT3 mutations during the course of disease in patients with acute myeloid leukemia

Abstract

NPM1 mutations have been reported to be the most frequent mutations in acute myeloid leukemia (AML). They are associated with a wide spectrum of morphologic subtypes of AML, normal karyotype and FLT3 mutations. The high frequency of NPM1 mutations might provide a suitable marker for monitoring residual disease of AML.

Nucleophosmin (NPM) is a multifunctional phosphoprotein with tumor-suppressor and oncogenic functions. NPM1 exon 12 mutations were found in 25–35% of adult de novo acute myeloid leukemia (AML). These mutations cause a frameshift and the formation of novel C-termini, and generate NPM mutants that localize aberrantly in the cytoplasm.1–4 Due to their frequency, NPM1 mutations may become a new tool for monitoring residual disease in AML.

We report on a comparison of the NPM1 and FLT3 mutational status during the clinical course of 28 adult AML patients. Bone marrow samples were collected from all patients after informed consent. The patients were diagnosed at our Institute and received induction-chemoterapy including standard dose Ara-C, Idarubicin and Etoposide, and consolidation therapy including Idarubicin and intermediate dose Ara-C. Patients with AML-M3 received All-trans-Retinoic Acid in addition to the chemotherapy described above. The presence of FLT3 mutations (Internal Tandem Duplication and point mutation at D835 residue) and NPM1 mutations were identified by the high sensitive Denaturing-High Performance Liquid Chromatography (D-HPLC) assay and direct sequencing, using the previously described primers for the FLT3 analysis,5 and the forward primer NPM1-F (5’GAAGAATTGCTTCCGGATGATC3’) and the reverse primer NPM1-R (5’CCTGGACAACATTTATCAAA-CACGGTA3’) for the amplification of NPM1 gene.

Mutations of NPM1 gene were present in 11/28 (39%) AML cases: type A mutation (960_963dupTCTG) occurred in 9/11 (82%) samples, type B (960_963ins CATG) (patient #2) and type D mutation (960_963ins CCTG) (patient #11) were each present in 1 case (9%).1 Overall, 12/28 patients (43%) carried a mutation of FLT3 at diagnosis (6 single ITD-mutations and 6 D835-mutations) (Table 1).

Table 1.Summary of patient characteristics and comparison of mutations during the clinical course.

We analyzed NPM1 and FLT3 mutations during the progression of disease, in a median follow-up of 11 months (range 3–31). The first relapse occurred at a median of 9.5 months (range 3–31) after diagnosis. Each patient had matched diagnostic and first relapse samples available for analysis. One of them also had a second relapse sample. Samples from 12 patients were also available at the time of first complete remission (CR1), and samples from 2 patients were obtained in second complete remission (CR2). CR and relapse were defined by classical morphologic criteria (i.e. less and more than 5% blasts in the bone marrow respectively). For the NPM1 mutation, we observed that the same mutation as that detected at diagnosis was identified again at first relapse in all NPM1-mutated patients. Furthermore, patient #1 also showed the same mutation at second relapse. No mutation in NPM1 was detected in relapses of patients that revealed wild-type NPM1 at diagnosis (17/28), as reported by others,2,4,6 although at relapse two of them showed a different karyotype from that at diagnosis (patients #17 and 18). In the samples obtained at the time of CR from patients harbouring NPM1 mutated at diagnosis, the mutation became undetectable. This shows that these were somatic mutations related only to the leukemic clone. Thus, in our experience, the NPM1 gene status was stable during disease. By contrast, we found that FLT3 mutational status changed between diagnosis and relapse in 7/28 patients (25%), 4 of them (patients #5, 9, 10, 11) also carried a NPM1 mutation. Two patients (#17 and 18) lost the mutation at relapse, 4 patients (#9, 10, 11, 19) acquired the mutation at relapse and patient 5 modified the mutation from D835 to ITD. In three patients, the change of FLT3 status was correlated to a modification of FAB or karyotype. In the patient 10, who also carried a NPM1 mutation, the FAB at diagnosis was M0, at relapse it evolved to M1 and acquired the FLT3-D835 mutations. Patient #17 acquired the t(9;19) and lost the FLT3-ITD at relapse. Patient #18, who harbored the t(9;22) and the FLT3-D835 mutation at diagnosis, lost both these alterations at relapse. Patients #1–2 and 12–14 relapsed with the identical FLT3-ITD length mutation types and patients #3–4 and 15–16 exhibited the same D835-mutation types at both stages. All samples obtained at the time of CR were negative for the presence of FLT3 mutations (Table 1).

In conclusion, NPM1 gene status was stable during disease evolution, in contrast to FLT3.6–8 The results reported suggest that NPM1 mutation, not FLT3 mutation may be considered as a potential marker for monitoring minimal residual disease. It could be useful to monitor residual disease in patients with normal karyotype, in which no alternative molecular markers are available. If the stability of NPM1 mutations at relapse is confirmed, these mutations may be useful to monitor residual disease in a large subgroup of AML patients.

Footnotes

• Funding: this work was supported by: COFIN 2005 (Myelodisplastic syndromes: pathogenetic models and promise of new therapies), COFIN 2003 (Molecular therapy of leukemias), FIRB 2001, University of Bologna, Italian Association for Cancer Research (A.I.R.C.), Italian National Research Council (C.N.R), Fondazione Del Monte of Bologna e Ravenna (Italy) and A.I.L. grants.