Increased Th17 and Treg levels in peripheral blood positively correlate with minimal residual disease in acute myeloid leukaemia

ABSTRACT

Purpose

Immune dysregulation plays a key role in acute myeloid leukemia (AML). We aimed to explore the correlation between T helper cell 17 (Th17) and the regulatory cells (Tregs) in the peripheral blood of patients with newly diagnosed (ND) AML and bone marrow blast cells, as well as minimal residual disease (MRD) before and after treatment.

Methods

Changes in Th17 and Treg cells in the peripheral blood of 32 patients with ND AML were observed before and after induction chemotherapy with cytarabine for seven days and anthracycline for three days. The levels of inflammatory cytokines were measured using an enzyme-linked immunosorbent assay. Correlation analysis between bone marrow blast cells and Th17 and Treg cell frequencies was performed using the Pearson’s correlation test. Frequencies of Th17 and Treg cells and MRD were assessed using flow cytometry.

Results

IL-6, IL-10, IL-17A, and GM-CSF levels gradually increased in patients with ND AML and CR and NR patients. The percentages of Th17 and Treg cells positively correlated with those of blast cells. In addition, the frequencies of Th17 and Treg cells in MRD-positive patients were higher than those in MRD-negative patients at the initial induction and after three months of chemotherapy. The frequencies of Tregs and Th17 cells positively correlated with MRD onset.

Conclusion

Increased Th17 and Treg cell levels were positively correlated with onset of AML, poor remission, and MRD.

KEYWORDS:

• Acute myeloid leukemia
• T helper cells
• regulatory cells
• Th-associated cytokines
• minimal residual disease

Introduction

Acute myeloid leukemia (AML) is a rare but incurable cancer, mostly induced by genetic abnormalities that lead to clonal proliferation and neoplastic changes in progenitors of the myeloid lineage [1,2]. Although chemotherapy can induce remission in patients with AML, limitations such as disease relapse, treatment resistance, and severe infections can still lead to high mortality [3,4]. The progression of AML and acquisition of drug resistance are associated with alterations in the bone marrow (BM) microenvironment, which offers physical protection for leukemia cells, releases pro-survival factors, induces dysregulated immune responses, and promotes the immune escape of AML blast cells [5]. Additionally, AML blast cells negatively affect various components of the immune system, mainly by weakening T cells. Notably, leukemia cells can dysregulate immune checkpoints and receptor-based signal transducers, leading to the negative regulation of T cells and ultimately evasion of immune surveillance [6]. Therefore, the immune system plays an indispensable role in the pathogenesis and treatment of this disease [7]. Consequently, a deeper understanding of immunity in AML is required for better prognoses and effective treatment.

The regulatory cells (Tregs) were originally identified as CD4⁺ CD25⁺ T cells [8]. Tregs are considered key regulators of immune evasion in tumor immunity, as they inhibit the proliferation and function of immune cells through cell-to-cell contact and inhibition of cytokine production. The number of Tregs increases in many hematopoietic malignancies [7,9]. Furthermore, increased Treg frequency is associated with a poor prognosis in pancreatic and breast cancer [10,11]. Tregs and T helper 17 (Th17) cells exert crucial effects on the immune homeostasis of the tumor microenvironment [12]. The levels of Th17-related cytokines, such as interleukin (IL)−23, transforming growth factor-beta, IL-1β, IL-6, IL-17, IL-22, and IL-21, differ between normal and malignant AML cells, suggesting that Th17 is a potential AML biomarker [13]. Therefore, we analyzed the ratio of Tregs to TH17 cells at different stages of AML. Furthermore, cytokines act as molecular messengers in the immune system, and an imbalance in their release has a remarkable effect on AML progression [14]. Nevertheless, the specific roles of these cytokines, Tregs, and Th17 cells in AML progression require further investigation.

Additionally, during the onset, treatment, and progression of AML, inadequate immune responses against certain myeloid leukemia blast cells may be associated with the onset of minimal residual disease (MRD) and subsequent relapse [15]. MRD can be used as a biomarker for making decisions regarding the routine care of patients with AML [16]. MRD negativity is emerging as a surrogate therapeutic endpoint that complements and may eventually replace the standard AML complete remission (CR) criteria [17]. A previous study has shown that tumor-associated antigen-specific cytotoxic T lymphocytes can prevent MRD positivity in AML [18]. However, the explicit roles of Tregs and Th17 cells and their interactions in the peripheral blood (PB) of patients with AML, as well as the relationship between Th17/Treg cells and MRD, remain unclear.

In the present study, the frequencies of Th17 and Treg cells, the associated cytokines in the PB of patients with AML, and the correlation between Th17/Treg and BM blast cells were explored. Additionally, the relationship between Th17/Treg cell frequency and MRD was investigated after the initial induction or three months of standard chemotherapy. Through this study, we aimed to reveal the role of Th17 and Treg cells in the diagnosis and remission of AML.

Methods

Patients and controls

The present study analyzed data from a prospective study that investigated the changes in Th17 and Treg cell levels in the PB of patients with AML. A total of 32 patients (18 men and 14 women; 36–65 years old; median age, 43 years) with newly diagnosed (ND) AML were enrolled in our hospital between November 2018 and December 2019. AML was diagnosed according to the French-American-British (FAB) classification system [19]. The exclusion criteria were as follows: (i) patients who had received chemotherapy before admission, (ii) patients with severe organ dysfunction or mental disorders, and (iii) patients who were in poor condition and could not tolerate chemotherapy. Based on the above exclusion criteria, the patients enrolled in the patient cohort of this study received subsequent chemotherapy. AML risk was evaluated according to the European Leukemia Net (ELN) risk stratification in 2017 [20]. CR was defined in patients with AML based on the International Working Group Criteria [21]. ND patients with AML underwent standard induction chemotherapy with cytarabine (100 mg/mq, intravenous injection) for seven days and anthracycline (doxorubicin 60 mg/m²) for three days. After ND patients with AML achieved CR, a high dose of cytarabine was administered as consolidation treatment. The specific treatment timelines for the patients are shown in Table S1. Patients demonstrating CR underwent MRD detection after three months (immune system recovery) of standard chemotherapy. A total of 24 individuals with mild iron deficiency anemia who underwent BM examination were selected as controls (14 women and 10 men; 17–67 years old; median age, 38 years) to exclude the influence of other hematological diseases on the indicators we subsequently detected. To rule out immunological perturbations, patients with fever not caused by AML and/or antinuclear antibody positivity were excluded. Cytogenetic analyses were performed centrally according to standard protocols, following the AML risk stratification of the ELN guidelines. Cytogenetic data were classified according to the International System for Human Cytogenetic Nomenclature [22,23]. Patients were classified into three subgroups based on cytogenetics: the group associated with a favorable prognosis included patients with AML with t (8;21), inv (16), or t (16;16). The adverse prognosis group included patients with AML with aberrations of chromosomes 5 or 7, aberrations of 11q23 or 17p, inv (3), t (3;3), or a complex aberrant karyotype (i.e. ≥3 clonal chromosomal aberrations). The group associated with an intermediate prognosis included patients with AML with other karyotype aberrations as well as AML with a normal karyotype. Written informed consent was obtained from all the participants in accordance with the Declaration of Helsinki. The study protocol was approved by the ethics committee of our hospital.

Sample collection

PB and BM blast cell samples were obtained from all participants after each therapy and stored in heparin-anticoagulant vacutainer tubes. PB mononuclear cells (PBMCs) were obtained from PB, and plasma was isolated from PB after centrifugation. Correlation analysis of BM blast cells with Th17 and Treg cell frequencies was performed using the Pearson’s correlation test. The timeline of sample collection is shown in Table S2.

Flow cytometric analysis for Th17 and Treg cell frequencies

The PBMCs were incubated with anti-CD4-APC and anti-CD25-PC7 antibodies (BD PharMingen, San Diego, CA, USA). After surface staining, the cells were fixed, permeabilized, and stained with anti-human IL-17APE and anti-human CD127-PC5 antibodies (BD Pharmingen) according to the manufacturer's instructions to perform intercellular staining. The samples were analyzed using flow cytometry. Specific materials and methods for the flow cytometric analysis of Th17 and Treg cell frequencies are available in the Supplementary Material.

Enzyme-linked immunosorbent assay (ELISA)

The concentrations of IL-6, IL-10, IL-17A, and granulocyte-macrophage colony stimulating factor (GM-CSF) were detected using ELISA kits (Esebio, Shanghai, China) according to the manufacturer’s instructions.

Flow cytometry analysis for MRD

MRD during the first course and three months after chemotherapy was detected using flow cytometry, as previously described [24]. The specific methods can be found in the Supplementary Materials. MRD > 0.1 was considered as high expression; MRD ranging from 0.001–0.1, medium expression; and MRD < 0.001, low expression. High and medium expression levels were defined as MRD + and low expression was defined as MRD-.

Statistical analysis

Data are expressed as mean ± standard deviation. SPSS software (version 13.0) was used to perform statistical analyses. One-way Analysis of Variance and Tukey's multiple comparison tests were used to evaluate differences among multiple groups. Multiple comparisons were performed using a two-way ANOVA and Sidak's multiple comparison test. Linear correlations were assessed using the Pearson correlation analysis. Each experiment was repeated three times. Statistical significance was set at P < 0.05.

Results

Analysis of patient characteristics

We analyzed the characteristics of ND patients with AML and controls. The demographic and clinical characteristics of the patients are presented in Table 1. The mean age of 32 ND patients with AML (men: 19 and women: 13) was 46.12 ± 3.20 years, whereas the mean age of 24 age-matched volunteers (men: 9 and women: 15) was 48.20 ± 2.82 years. The mean lymphocyte count was decreased in ND AML group (1.83 ± 1.02) than that in the control group (2.16 ± 0.70). The mean white blood cell count was relatively higher in the ND AML group (6.31 ± 10.08) than that in the control group (5.27 ± 1.24). The percentage of BM leukemic blast cells in the ND AML group was 62.21 ± 2.68, whereas it was only 0.19 ± 0.06 in the control group. A total of 32 ND patients with AML were characterized based on the FAB classification. No patients showed the M1 subtype, whereas 15, 6, and 11 patients showed the M2, M4, and M5 subtypes, respectively. In terms of cytogenetic-molecular risk stratification, favorable (n = 12), intermediate (n = 16), and adverse (n = 4) outcomes were recorded. Additionally, six, one, and two patients had NPM1, FLT3, and CEBPA mutations, respectively. Among the 32 ND patients with AML, 29 achieved CR, of whom 24 achieved CR after one course of chemotherapy, whereas three abandoned treatments.

Table 1. Baseline data in ND patients with AML and controls.

Features	ND AML patients (n = 32)	Controls (n = 24)
Age (years, mean ± s.d.)	46.12 ± 3.20	48.20 ± 2.82
Gender (male/female)	19/13	9/15
lymphocyte (*10^9/L)	1.83 ± 1.02	2.16 ± 0.70
WBC (*10^9/L) BM leukemic blast (%) FAB subtype M1 M2 M4 M5 cytogenetic-molecular risk stratification Favorable Intermediate Adverse Mutation status NPM1 FLT3 CEBPA	6.31 ± 10.08 62.21 ± 2.68 0 15 6 11 12 16 4 6 1 2	5.27 ± 1.24 0.19 ± 0.06

WBC: white blood cell; BM: bone marrow; FAB: French-American-British; NPM1: nucleophosmin; FLT3: FMS-like tyrosine kinase-3; CEBPA: CCAAT/enhancer binding protein alpha.

Treg/Th17 cell ratio in PB of patients with AML is downregulated after treatment

Before and after treatment, the frequencies of Th17 and Treg cells in different phases of AML were determined using flow cytometry. After first standard chemotherapy, the frequencies of PB Th17 (3.31 ± 0.31%) and PB Treg (4.79 ± 1.4%) cells in patients with AML demonstrating CR were both higher than those of ND patients with AML (Figure 1(A,B), P < 0.05). In addition, both PB Th17 (4.26 ± 0.46%) and PB Treg (5.51 ± 0.86%) cell frequencies in patients with AML showing non-CR (NR) were higher than those in patients with AML showing CR (Figure 1(A,B), P < 0.01). Next, the Treg/Th17 cell ratio in the different phases of AML was investigated. After the first induction treatment, a decreased Treg/Th17 cell ratio was observed in patients with AML showing CR than that in ND patients with AML (P < 0.01). We observed a higher Treg/Th17 cell ratio in patients with AML showing NR than in those with AML showing CR (P < 0.05; Figure 1(C)).

Figure 1. Th17/Treg cell frequency and Treg/Th17 ratio in the PB of patients with AML at different phases (A) The frequency of PB Th17 cells in different phases of AML was analyzed by flow cytometry. **P < 0.01 vs. the CR, NR or control group. (B) The frequency of PB Treg cells in different phases of AML was analyzed by flow cytometry. *P < 0.05 vs. the CR group, **P < 0.01 vs. the NR or control group. (C) The ratio of Treg/Th17 in different phases of AML. *P < 0.05 vs. the NR group.

Levels of T cell-associated cytokines in patients with AML are elevated after treatment

To understand the influence of T-cell-associated cytokines on AML prognosis, we determined the plasma levels of these cytokines after standard induction chemotherapy. A significant increase in plasma levels of IL-6, IL-10, IL-17A, and GM-CSF (17.13 ± 1.8, 6.32 ± 0.73, 4.08 ± 0.37, and 13.1 ± 1.13 pg/mL, respectively) was observed in patients with AML showing CR than those in ND patients with AML. The plasma levels of IL-6, IL-10, IL-17A, and GM-CSF were all significantly lower in the CR group than in the NR group; however, these levels were still higher than those in the control group (Figure 2(A), P < 0.01). Additionally, as shown in Figure 2(B,C), significant positive correlations were found between IL-17A and IL-6 levels (P = 0.008, r = 0.4604) and between IL-17A and GM-CSF levels (P = 0.0009, r = 0.5569) in PB plasma after treatment. These results indicate that T cell-associated cytokines were increased in ND patients with AML, which was in line with the changes in the frequencies of Treg and Th17 cells. We hypothesized that these cytokines would be associated with the prognosis of patients with AML.

Figure 2. T cells-associated cytokines in different phases of AML. (A) The concentrations of cytokines in PB of AML before and after treatment. **P < 0.01 vs. the control group, ^##P < 0.01 vs. the ND group, ^&&P < 0.01 vs. the NR group. (B) IL-6 concentration was positively correlated with IL-17A concentration. P = 0.008, r = 0.4604. (C) GM-CSF concentration was positively correlated with IL-17A concentration. P = 0.0009, r = 0.5569.

Treg and Th17 cell frequencies are reduced in patients with MRD after three months of standard chemotherapy

The MRD levels in patients with CR were detected using flow cytometry. After initial treatment, CR was achieved in 24 patients (data not shown). Among them, 18 patients (75%) were MRD-negative (MRD^-) and six (25%) were MRD-positive (MRD⁺). As demonstrated in Figure 3(A,B), both Th17 and Treg cell frequencies were lower in the initial induction MRD ^– group than in the initial induction MRD⁺ group (P < 0.01). After three months of standard chemotherapy, 26 patients achieved continuous CR and three patients experienced recurrence. A total of 23 of the 26 CR patients (88.5%) were MRD^-, and three patients (11.5%) were MRD ⁺. In addition, the frequencies of Th17 and Treg cells were lower in the three-month MRD ^– group than in the three-month MRD⁺ group (P < 0.05). Furthermore, Th17 and Treg frequencies were relatively lower after three months of treatment than after initial treatment (Figure 3(A,B)).

Figure 3. Treg and Th17 frequencies are reduced in MRD after three months of standard chemotherapy. (A) The frequency of PB Th17 cells was decreased in MRD negative patients compared to MRD positive patients after the initial standard induction or three months of standard chemotherapy. **P < 0.01 vs. the initial induction MRD⁺ group or 3 months MRD⁺ group. (B) The frequency of PB Treg cells was decreased in MRD negative patients compared to MRD positive patients after the initial standard induction or three months of standard chemotherapy. *P < 0.05 vs. the 3 months MRD⁺ group, **P < 0.01 vs. the initial induction MRD⁺ group.

Treg and Th17 cell frequencies are positively correlated with high-risk stratification

We further analyzed the association between Th17/Treg cells and BM blast cell counts in ND patients with AML. As shown in Figure 4(A,B), positive correlations were observed between BM blast cell counts and Th17 (P < 0.01, R²= 0.3912) and Treg cell counts (P < 0.01, R²= 0.4502). In addition, the frequencies of Th17 and Treg cells in PB were significantly higher in the high-risk group than in the intermediate – or low-risk groups (Figure 4(C), P < 0.01).

Figure 4. Treg and Th17 frequencies are positively correlated with high-risk stratification. (A) Positive correlation between PB Th17 cell frequency and bone marrow blast cells. P < 0.01, R² = 0.3912. (B) Positive correlation between PB Treg cell frequency and bone marrow blast cells. P < 0.01, R² = 0.4502. (C) The relationship between T cell subsets and risk stratification. **P < 0.01 vs. the high-risk group.

Discussion

CR is achieved in 60% of ND patients with AML receiving first-line therapy; however, relapse has been reported in 30–40% of patients [25]. In addition to cytogenetic and molecular changes, immune escape also plays an important role in AML pathogenesis [26,27]. In the present study, we investigated Treg and Th17 cell levels and cytokine production in patients with AML before and after treatment, and their correlation with MRD. Increased Th17 and Treg cell levels were positively correlated with AML onset, poor remission, and MRD.

An imbalance in the Th17/Treg cell ratio is involved in AML pathogenesis. Flow cytometry-based analysis of PB samples from patients with AML and immune-related risk factors demonstrated immune dysfunction of lymphocytes, suggesting that immune signatures can predict therapeutic reactivity, high risk of relapse, and unfavorable prognosis [28]. Significant differences were observed between normal and AML BM immune cells and increased frequencies of TH17 and Treg cell clusters [27]. In this study, the frequencies of Th17 and Treg cells were significantly increased in the PB of patients with AML compared to those in controls, consistent with previous results [29–31]. Importantly, after treatment, the Treg/Th17 cell ratio decreased in the CR group compared with that in the ND and NR groups. These results suggested that an imbalance in the Treg/Th17 cell ratio contributes to the regulation of immune function in patients with AML. However, the frequencies of Th17 and Treg cells were higher in CR patients with AML than in ND patients with AML. This phenomenon may be attributed to the fact that immune function is temporarily disturbed by the toxic effects of chemotherapy drugs, which are not fully recovered at that time. Moreover, patients with higher frequencies of Th17 cells have lower survival rates than those with lower frequencies [32,33]. Our results demonstrated that the frequencies of Th17 and Treg cells in PB were positively correlated with risk stratification. We believe that the reason for this contradiction is that Th17 cell frequencies increase persistently when CR is achieved [34]. However, in contrast to Th17 cells, relatively high Treg cell frequencies are maintained, which no longer demonstrate an increase when CR is achieved [35]. Additionally, a positive correlation was observed between T cell subsets and high-risk stratification. These results suggest that Tregs and Th17 cells are potential immunotherapeutic targets.

In this study, we found increased levels of IL-6, IL-10, IL-17A, and GM-CSF in the ND, CR, and NR groups of patients with AML than in the controls. Whereas, the levels of IL-6, IL-10, IL-17A, and GM-CSF were lower in the CR group than in the NR group. The data demonstrated a positive correlation between IL-17 and IL-6 levels and IL-17A and GM-CSF levels. One study reported that the plasma levels of IL-6, IL-17A, and IL-10 were higher in patients with AML [36]. Additionally, GM-CSF plays important roles in the proliferation, differentiation, and survival of myeloid cells [37]. Th17 cells are a group of pro-inflammatory CD4+ effector T cells that produce IL-17A [38]. Based on the aforementioned results, the levels of cytokines associated with Tregs and Th17 cells increased in ND patients with AML, in line with the changes in Tregs and Th17 cells. A significant increase in plasma IL-6 levels has been observed in ND patients compared to controls, and IL-17A is positively correlated with IL-6 concentration [39]. Our findings indicated that T cell-related cytokines may be involved in AML pathogenesis.

Accumulating evidence has demonstrated that MRD is an effective indicator of AML prognosis [40,41]. MRD negativity is associated with superior disease-free and overall survival in patients with AML [42]. In our study, 75% of patients who were MRD-negative showed decreased frequencies of Th17 and Treg cells after the initial standard induction treatment. Furthermore, 88.5% of the CR patients were MRD-negative, and the percentages of Tregs and Th17 cells were lower than those in the positive group after three months of standard chemotherapy. Our results further suggest that the immune system of these patients may gradually recover after three months of chemotherapy. Similarly, a correlation has been observed between the number of Tregs in the BM and MRD [43]. These findings may help develop strategies and provide insights into the post-treatment management of patients with AML.

This study had some limitations. Currently, AML is diagnosed and classified using the World Health Organization system, which is newer than the FAB classification used in our study. Additionally, the sample size was relatively small, and a larger sample size was required. The methodology for T cell detection and blood sampling time remains to be optimized. Furthermore, IL-6, IL-10, and GM-CSF are not specific T cell cytokines, and we are considering the use of specific cytokines in future studies. This study was limited to patients receiving the 3 + 7 regimen, and further investigations are needed in patients receiving other chemotherapeutic drugs. Finally, as we had limited samples and experimental conditions at this stage, we were unable to analyze the mutation status of more genes or use multiparameter flow cytometry with more color panels. We intend to refine this research in future studies.

Taken together, the increased frequency of Th17 and Treg cells and elevated T cell-associated cytokine (IL-6, IL-10, IL-17A, and GM-CSF) levels were strongly associated with AML onset, presence of MRD, and poor remission. The frequencies of Th17 and Treg cells in PB were positively correlated with risk stratification. Additionally, the levels of cytokines associated with Tregs and Th17 cells increased in ND patients with AML, indicating that T cell-related cytokines may be involved in AML development. Increased frequencies of Th17 and Treg cells are promising markers of MRD positivity in AML. We hope that our findings will provide a theoretical basis for further improvements in the prevention, diagnosis, and treatment of AML.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by Shandong Province Medical and Health Technology Development Plan Project [grant number 202003041248].