AML-8: Clinical Meaning and Prognostic Implications of Non-Canonical Non-Type A NPM1 Mutations in Acute Myeloid Leukemia

Challenge

NPM1 mutated variants of acute myeloid leukemia (AML) account for around one-third of all de novo AML. Type A accounts for about 80% of all NPM1 mutations, and other rarer mutations with diverse insertions account for the remaining 20%. Type A can be classified as favorable or intermediate risk based on either the absence or presence of FLT3 mutations, respectively, in the absence of adverse cytogenetic features. However, for non-type A mutations, too little data of sufficient quality is available to make such a risk prognosis. Published studies have a limited cohort size, mainly investigated just the most common non-A NPM1 mutations and, in some cases, failed to consider prognostic factors in NPM1-mutated AML, such as the FLT3 status. This lack of an accurate risk prognosis for AML patients with a non-type A NPM1 mutation hampers clinical decisions regarding the most appropriate treatment strategy. 

---

Project Partnership

• University of Milan-Bicocca, Italy
• IRCCS San Gerardo dei Tintori, Italy
• Policlinico Hospital Milan, Italy

---

Project Leadership

• Valentina Sangiorgio, IRCCS San Gerardo dei Tintori Monza and IRCCS Candiolo, Italy

---

Impact

• A more accurate stratification of the prognostic impact of non-type A NPM1 mutated AML variants could lead to a better allocation of individual AML patients in the appropriate risk category.
• This should increase the chance of each of these patients receiving the most appropriate treatment strategy, including the need to consolidate disease remission with an allogeneic bone marrow transplant.

---

Project Summary

For AML patients with a non-type A NPM1 mutation, too little data is available to make an accurate risk prognosis, and that hampers clinical decisions regarding the most appropriate treatment strategy. Therefore, this study aims to investigate the real prognostic impact of rare non-type A NPM1 mutations and correlate the different synonymous variants with known prognostic factors to make a more accurate risk prognosis for individual patients possible.

This study started with 75 NPM1 mutated AML cases from the hospital and was subsequently expanded to a cohort of 550 cases by using well-known and publicly available datasets. This cohort was divided into NPM1-A mutated versus NPM1-non A mutated. The non-A group was further analyzed to couple all synonymous variants and key clinical variables were collected within each distinct variant group. This real-world cohort is already powered to produce a reliable analysis of the impact of non-A variants, but increasing its size with the use of HARMONY data will make the analysis more robust.

The overall objectives are: 

• Increase understanding of the clinical and prognostic meaning of rare non-type A NPM1 variants
• Better stratify the prognostic impact of non-type A NPM1 variants.
• Improve the allocation of individual non-type A patients to the appropriate risk category.

This study should result in a better risk prognosis for AML patients with non-A type NPM1 variants and that will facilitate them getting the most appropriate treatment. We speculate that some of these variants may have a more aggressive behavior (perhaps regardless of the FLT3 status) and patients with these variants may benefit from an allo-BMT approach upfront. On the contrary, other variants may harbor a more indolent impact and patients with these might benefit from additional target therapies (such as Gemtuzumab Ozogamicin) rather than an allo-BMT9. 

---