Mesenchymal Stem Cells in NPM1 Mutated Low Risk Acute Myeloid Leukemia: a Study of the Tumor Microenvironment and Its Contribution to the Outcome

Background Acute myeloid leukemia (AML) is a heterogeneous hematological malignancy characterized by clonal expansion and differentiation arrest of myeloid progenitor cells.

Although incomplete, current knowledge of AML pathogenesis has allowed patient stratification into different prognostic groups and has led to the emergence of personalized treatment strategies.

The European Leukemia Net (ELN) stratifies de novo AML into favorable, intermediate and adverse risk groups according to genetic mutations and chromosomal abnormalities. In particular, patients with a favorable-risk profile are characterized by either the presence of core-binding factor rearrangements (CBF), normal karyotype with nucleophosmin 1 (NPM1) mutations and fms related receptor tyrosine kinase 3 (FLT3) wild-type, NPM1 mutations and FLT3-internal tandem duplication with low allelic ratio, or biallelic mutations in CCAAT/enhancer binding protein α (CEBPα).

Despite the favorable-risk profile, patients in long-term follow-up after conventional chemotherapy have relapse rates of up to 50% in AML with NPM1 mutation, 40% in CBF AML, and 44% in AML with CEBPα biallelic mutation. These data highlight the need for further investigation among favorable-risk AML to gain more knowledge regarding the pathogenetic and resistance mechanisms, thus improving patient stratification. Moreover, these insights could lead to the discovery of new therapeutic targets.

NPM1 is the most frequent mutation in AML, accounting for approximately 50-60% of de novo cytogenetically normal AML. Murine and in vivo studies have demonstrated that NPM1 mutation promotes leukemogenesis and myeloproliferation but is not sufficient to induce leukemia. Pre-existing mutations of hematopoietic stem cells (HSCs) may be a possible mechanism that cooperates in leukemogenesis. These mutations mainly involve epigenetic modifiers in hematopoietic stem or progenitor cells, resulting in their clonal expansion in individuals without an overt hematologic disease. This condition, named clonal hematopoiesis, represents a pre-leukemic state that may promote the emergence of mutations able to drive leukemogenesis. Indeed, NPM1 mutations frequently cluster with co-occurring mutations involving genes of the signaling pathways, DNA methylation, RNA-splicing, cohesin complex, gene transcription, and tumor suppression.

Although few studies have shown that some co-occurring mutations negatively affect disease prognosis, controversies remain on their real impact. Therefore, the mutational landscape alone cannot fully explain the clinical courses of some patients in this group of low risk leukemias; other factors may play a role in the heterogeneity of this AML subgroup. Recently, major efforts have been made to gain insights into the leukemic hematopoietic niche, which is implicated in disease refractoriness and relapse.

Haematopoietic Stem Cells regulate their proliferation, differentiation and self-renewal processes in a specific bone marrow niche, through the interaction with the BM microenvironment. The microenvironment has a cellular compartment that consists mainly of mesenchymal stem cells (MSCs), but also osteoblasts, endothelial cells, macrophages. In particular, MSC subgroups have different roles in the regulation of HSCs activity.

Leukemia stem cells are a small subgroup of leukemic cells capable of self-renewal and with stemness properties. By definition, these cells can re-initiate leukemia if engrafted in a recipient host.

The BM microenvironment interacts, as well, with leukemic stem cells (LSCs), altering their proliferation, adhesion properties, quiescence and clonal expansion.

At the same time, studies have shown that LSCs can reprogram the MSCs and receive trophic support, protection from oxidative stress and preservation from toxic chemotherapeutic agents, promoting the formation of a "leukemic niche". On the other hand, the BM MSCs interact with LSCs, directly or indirectly, and alter LSCs' proliferation, adhesion properties, quiescence and clonal expansion. As a result, the MSCs contribute to the LSC maintenance. MSC remodeling is a leukemogenesis selfr-enewal strategy that could explain the clinical heterogeneity and thus could be a potential prognostic factor.

Several studies have revealed the presence of functional alterations at the BM-MSCs level in patients with AML, such as decreased proliferation, increased apoptotic rate, and a pro-inflammatory state.

Although the BM microenvironment seems to play a key role in promoting chemoresistance in AML, whether these alterations or others are related to leukemia pathogenesis still remains unclear.

Patients with AML have altered MSCs characteristics; in particular, there is a reduced amount of primitive MSCs, defined as CD146+/nestin+, compared to healthy donors (HD). Nestin+ BM-MSCs enhance leukemic cell survival and chemoresistance by stimulating oxidative phosphorylation and thus increasing energy production. This subpopulation seems to be associated with the disease course: patients who relapse have an increased percentage of CD146+/nestin+ MSCs with respect to subjects in remission.

Moreover, modified MSCs interact differently with HSC and LSC, selectively inhibiting normal hematopoiesis and promoting leukemogenesis and chemoresistance, respectively. The mechanism of stroma-mediated protection of leukemic cells is complex and involves a network of cytokines, chemokines, soluble factors and microvesicles (MVs) that promote LSCs proliferation. A probable mechanism of action is the interaction between BM stromal cells and LSCs through the release of CXCL12 by MSCs and its binding to the CXCR4 receptor on the surface of LSCs. More in detail, leukemic cells exploit the CXCL12/CXCR4 axis for homing in a protective BM niche, which is usually restricted to normal HSC. As a result, LSCs dwell in an environment that promotes their survival and protects them from chemotherapy. For this reason CXCL12/CXCR4 axis could be targeted by future therapies to prevent relapses.

To our knowledge, there is no data regarding the characterization of the BM microenvironment, in NPM1mutated AML. The characterization of MSCs, as the most representative components of the BM microenvironment, could be useful for the identification of specific cellular mechanisms that might explain the high relapse rate observed in these patients.

Objectives of the study

General objective:

To characterize the BM-MSCs of subjects affected by AML NPM1mut as possible novel indicators of patient clinical outcome.

In detail, it will be studied the AML-MSCs' phenotype, genetic profile and function, and evaluate their paracrine activity in order to understand the mechanisms that promote resistance to therapy and relapse occurrence in AML NPM1mut at the level of the bone marrow microenvironment. The paracrine activity of MSCs' will be assessed through the study and characterization of the microvesicles released by MSCs after deprivation.

Primary Objective:

The primary objective of this study is the detection of CD146+ MSCs percentage at diagnosis, as this subpopulation seems to be involved in the LSC niche creation and chemo-resistance. It has been described that BM of relapsed patients shows higher percentage of this MSC subpopulation compared to patients undergoing clinical remission. The aim is the demonstration that in AML NPM1mut patients there is a statistically significant difference in CD146+ MSC percentage (at diagnosis) compared to HD-MSCs and DC-MSCs (disease control MSCs: AML patients NPM1 wt).

Co-primary objective:

Comparison of the percentage of CD146+ cells at diagnosis and at different time points (diagnosis, post induction, end of treatment, follow up, pre-transplant, relapse) between relapsed patients and patients who do not relapse in the whole population (AML NPM1mut and AML-NPM1wt)

Secondary objectives:

1. Compare the MSCs CD146+ subpopulation in AML NPM1mut and AML NPM1wt patients at each predefined time-point.
2. Characterize AML-MSCs in patients with AML NPM1mut vs AML NPM1wt. Following the international guidelines for the definition of MSCs (Dominici et al., 2006), will be evaluated the morphology, the ability to adhere to plastic, the clonogenic efficiency, the proliferative capacity and the transition to senescence. The expression of surface markers CD105, CD73 and CD90 and the lack of CD45, CD34, CD14, CD19 and HLA-DR will be evaluated through labeling with monoclonal antibodies and assessment by means of multiparametric flow cytometry (MFC) (FacsCanto II, Becton Dickinson). CD146 and nestin expression will be assessed on the samples obtained at specific time points. The AML-MSCs' ability to differentiate into adipocytes and osteocytes will be determined after "in vitro" stimulation with the appropriate differentiation factors.
3. Functional characterization of MSCs from AML NPM1mut vs AML NPM1wt patients. MSCs' functional properties will be evaluated through the development of a co-culture of AML-MSCs in the autologous setting (patient's peripheral blood mononuclear cells (PBMCs)/patient's MSCs) and allogeneic one (healthy donor's PBMCs/patient's MSCs). Then it will assessed the immunomodulating potential, the ability to generate T, B and NK subpopulations and the regulation of the CXCR4/CXCL12 axis. Moreover, the pro-inflammatory and regulatory cytokines in the culture's supernatant will be quantified.
4. Genomic characterization of the MSCs in patients with AML NPM1mut vs AML-NPM1wt By using next generaton sequencing (NGS) with the aid of a panel of genes recurrently mutated in myeloid neoplasms.

Tertiary explorative objective

• Sorting of the CD146+/nestin+ MSCs subpopulation for future co-culture in vitro studies
• Isolation and characterization of the MVs released by AML-MSCs after deprivation.
• Development of cultures of PBMCs from HD with MVs in order to evaluate their paracrine activity.

Study design This is a pilot single center prospective observational, longitudinal, in vitro study

Subjects and inclusion criteria Adult patients with new onset Acute Myeloid Leukemia diagnosed at Clinica Ematologica of the Fondazione IRCCS Policlinico San Matteo di Pavia.

The controls are healthy donors of haematopoietic stem cells for bone marrow transplant enrolled by Oncoematologia Pediatrica, upon signing the informed consent

Collection of Samples Expansion and analysis of the bone marrow MSCs at selected time points (diagnosis, end of induction, end of consolidation, follow-up, before hematopoietic stem cells transplantation (HSCT), relapse).

At the same time-points, a peripheral blood sample will be collected. Each planned time points represents an ordinary clinical practice bone marrow aspirate and blood collection.

Methods

Primary and co-primary objectives:

Sixty patients with new onset Acute Myeloid Leukemia diagnosed at Clinica Ematologica of the Fondazione IRCCS Policlinico San Matteo of Pavia will be enrolled in three years;

Primary and co-primary objective:

As standard procedure, the bone marrow aspirate of enrolled patients will be collected at diagnosis end of induction, end of consolidation, follow-up, before and after HSCT, relapse.

The MSCs will be isolated from the bone marrow of AML NPM1mut and AML-NPM1wt subjects, and than cultured "in vitro'' with 5% lysed DMEM+ culture medium until senescence.

The expression of the surface markers CD146 and nestin will be evaluated through intracytoplasmic labeling and subsequent analysis through MFC FacsCanto II (BD). The percentage of CD146+/nestin+ MSCs at disease onset in patients with AML NPM1mut with the percentage of HD and disease controls will be compared.

Secondary objectives:

The characterization of patients' BM-MSCs by evaluating the clonogenic efficiency (colony forming units, CFU-f). Meanwhile, the proliferative capacity will be assessed in terms of cumulative population doubling (cPD). Senescence of AML-MSCs will be defined as the time when the number of cells recovered is less or equal than the number of cells cultured. Senescent cells will be detected by using β-galactosidase.

The ability of AML-MSCs to differentiate into adipocytes and osteocytes, driven by the addition of specific factors in vitro: the Alizarin red dye specifically colors lipid droplets, whereas the alkaline phosphatase dye will highlight the presence of calcium deposits, typical of osteocytes.

At the same time-points, peripheral blood sample will be collected, a residue of the standard diagnostic analyses performed during clinical activity. From the sample PBMCs will be extracted and used in the subsequent immunomodulation experiments.

AML-MSCs' functional properties will be evaluated after co-culture with PBMCs. After three days of incubation, MSCs' ability to modulate PBMCs' proliferation will be analysed through incorporation of Titrated Thymidine. The results will be compared with those obtained from HD-MSC and disease controls.

After ten days of incubation in the same culture, the capacity of AML-MSCs of regulating the percentage of the lymphocyte subpopulations (T and B lymphocytes and NK cells) through MFC will be evaluated, along with the percentage of T regulatory cells through the measurement of intracytoplasmic Foxp3, and the expression of CXCR4 on the surface of patients' lymphocytes. The culture's supernatants will be stored and used to dose CXCL12, IL-6, IFN-γ, IL-10, PGE-2 by means of ELISA.

The DNA extracted from the MSCs of subjects with AML NPM1mut and AML NPM1wt will be analyzed through a custom TruSight Myeloid Sequencing Panel of 54 genes sequenced by Illumina MiSeq, together with DNA extracted from leukemic blasts of the same patients.

Tertiary explorative objective If the percentage CD146+/nestin+ MSCs will be suitable, this subpopulation will be sorted and the above-mentioned analysis on it will be performed.

The MVs will be isolated through ultracentrifugation and quantified (NanoSight). The expression of MVs surface markers will be assessed through MACSPlex Exosomes Kit (Miltenyi) and MacsQuant Flow Cytometer (Miltenyi).

The protein content of the MVs will be evaluated through HPLC and mass spectrometry.

Co-culture experiments in the presence of MVs will be developed.

Data Collection Data will be recorded in an electronic database. Patients will receive an identification code that ensures their privacy. The PI will keep separately a list with patient's identification data and the corresponding code.

Sample size calculation Ninenty subjects will be enrolled (60 patinents: 30 AML NPM1mut, 30 AML NPM1wt and 30 HD-MSCs). With this sample size it will be possible to detect a significant 2x2 post-hoc comparison (two-sided type I error alpha=0.017) in nestin+/CD146+ MSCs levels after one-way analysis of variance if the difference between means is 85% of the common standard deviation. A power of 80% has been considered.

Statistical analysis Qualitative variables will be described as absolute and relative frequencies of each category. Quantitative variables will be summarized as mean±standard deviation, or by median and interquartile range. If necessary, log-transformations of data will be adopted.

Statistical analyses will be performed using Stata 17 (StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC.)

Statistical analysis for the primary endpoint Nestin+/CD146+ MSCs levels at baseline will be compared between three groups with one-way analysis of variance, followed by Bonferroni's 2x2 post-hoc comparisons.

Statistical analysis for the co-primary endpoint The association between nestin+/CD146+ MSCs levels at different time points and relapse will be evaluated by a joint model of longitudinal and survival data.

Statistical analysis for the secondary endpoints

1. The comparison according to NPM1 mutational status will be carried out fitting multilevel mixed (with patients treated as random effect) linear or logistic models.
2. The interactions between MSCs and leukemic cells will be analyzed by fitting multilevel mixed (with patients treated as random effect) linear or logistic models.
3. Fisher's exact test will be performed in order to compare patients with AML NPM1mut vs AML-NPM1wt.

Statistical analysis for tertiary explorative endpoints For tertiary explorative endpoints, descriptive statistics will be employed.

Informed consent Each patient will receive the study informative material and signed informed consent form will be obtained.

An information leaflet and informed consent will clarify that the patient will not be subjected to additional blood withdrawals and bone marrow aspirates.