Targeting CD33 for acute myeloid leukemia therapy

Background The aim of this study was to analyze the level of CD33 expression in patients with newly diagnosed AML and determine its correlation with clinical characteristics. Methods Samples were collected for analysis from AML patients at diagnosis. We evaluated the level of CD33 expression by flow cytometry analysis of bone marrow. Chi-square or t- tests were used to assess the association between the high and low CD33 expression groups. Survival curves were generated by the Kaplan-Meier and Cox regression model method. Results In this study we evaluated the level of CD33 expression in de novo patients diagnosed from November 2013 until January 2019. The mean value of 73.4% was used as the cutoff for the two groups. Statistical analysis revealed that 53 of the 86 (61.2%) AML patients were above the mean. Although there was no statistical significance between CD33 expression level and gene mutation, FLT3 mutation (P = 0.002) and NPM1 mutation (P = 0.001) were more likely to be seen in the high CD33 group. The overall survival (OS) was worse in the high CD33 group (39.0 m vs. 16.7 m, x2 = 13.06, P < 0.001). The Cox survival regression display that the CD33 is independent prognostic marker (HR =0.233,p = 0.008). Univariate analysis showed that the high expression of CD33 was an unfavorable prognostic factor. Of the 86 patients, CD33-high was closely related to the patients with normal karyotype (x2 = 4.891,P = 0.027), high white blood cell count (WBC, t = 2.804, P = 0.007), and a high ratio of primitive cells (t = 2.851, P = 0.005). Conclusions These findings provide a strong rationale for targeting CD33 in combination with chemotherapy, which can be considered a promising therapeutic strategy for AML. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-021-09116-5.


Introduction
In adult patients, acute myeloid leukemia (AML) is a common hematological malignancy. Most of the patients have a poor prognosis. Conventional chemotherapy for AML, including induction and consolidation treatment, is only partially effective. Patients often require bone marrow transplantation and multiple rounds of consolidation therapy. Even with regular chemotherapy, the overall 5-year survival remains below 30% for all patients and is lower for older patients. Most importantly, many patients are refractory to conventional chemotherapy [1][2]. Thus, we need to develop new methods to improve survival in this fatal form of leukemia. Recently, immunotherapy has been recognized as a new treatment strategy for hematologic malignancies [3][4]. CAR-T immunotherapy has shown excellent results in ALL (acute lymphocytic leukemia) and lymphoma, and it is possible that patients with acute myeloid leukemia (AML) can also bene t from this treatment strategy [5].
Unlike ALL, AML surface antigens are highly heterogeneous. We need to identify antigens that are speci cally expressed in AML as markers for immunotherapy. CD123, CD44, CD174 and CD33 are expressed in hematopoietic stem and progenitor cells, increasing off-target toxicity and killing hematopoietic stem and progenitor cells in immunotherapy [3,[6][7].
Most efforts of developing monoclonal antibodies or antibody-drug conjugates (ADCs) for AML have focused on targeting CD33 (cluster of differentiation antigen 33). Leukemic blasts and myeloid leukemiainitiating cells express CD33. CD33 does not appear on the surface of primitive stem cells or multipotent progenitor cells. These factors make it a favorable target for immunotherapy of AML [1,[8][9][10]. There have been a number of reports con rming that CD33 is a feasible target for immunotherapy of AML. Due to the approval of anti-CD33 Mylotarg® (GO, gemtuzumab ozogamicin) in 2000, GO was the rst anticancer ADC on the market[8, [10][11][12]. In our study, we investigated the correlation between the level of CD33 expression in patients with newly diagnosed AML and the prognosis of patients. All data were obtained from 86 newly diagnosed AML patients. Of course, all processes met the ethical standards, and patient consent was obtained. This study provides more persuasive evidence for the immunotherapy of AML.

Patients And Method
Patients Between November 2013 and January 2019 in our institution (Department of Hematology, First A liated Hospital of Henan University of Science and Technology), 86 patients with an initial diagnosis and complete information were enrolled in the study group. All patients were diagnosed, evaluated and treated according to National Comprehensive Cancer Network (NCCN) guidelines [13]. All patients' records were evaluated retrospectively for the level of CD33 expression in de novo AML patients. Patients were grouped according to expression levels above and under the mean, that is, into high and low level of CD33 expression groups. The association between patient clinical information and CD33 expression was analyzed. Detailed baseline characteristics are shown in Table 1. Table 1 Patients' characteristics at initial diagnosis. Treatments Induction chemotherapy used standard DA (daunorubicin + cytarabine), IA (nordomycin + cytarabine), MA (mitoxantrone + cytarabine) or CAG with or without the D (arabin cytidine + aclamycin + granulocyte colony stimulating factor with or without decitabine) regimen. Consolidation chemotherapy was given after complete remission (CR), and the regimen referred to induction chemotherapy and usually included cytarabine. For patients with moderate or poor prognosis, allogeneic hematopoietic stem cell transplantation should be performed with appropriate, medium and large dose regimens.

Methods
A multicolor immunolabeling method was used to determine the expression ratio of CD33. The proportion of CD33 expression was determined by FACS analysis (FACS Navios, Beckman Coulter). as described [14][15]. Bone marrow chromosome karyotype analysis was performed by the R-banding method. Real-time quantitative PCR (RT-PCR) assay was used to detect fusion genes. Gene mutations The chi-square test (2-tailed) was used when comparing the categorical variables; a value of p ≤ 0.05 was considered statistically signi cant. Fisher's exact test and t-test were used under the right conditions.
The Kaplan-Meier method was used to construct survival curves. Overall survival was calculated from the time of diagnosis to all-cause death. In univariate and multivariate analyses, log-rank tests were applied to analyze differences in OS between groups. Estimation was limited to the largest survival if it was censored.

Results
Patient characteristics. Table 1 summarizes all the characteristics at diagnosis of the 86 patients in the retrospective analysis in the study. Patients were divided into high level of CD33 expression (High-CD33) and low level of CD33 expression (Low-CD33) groups. We found signi cant differences in the WBC (white blood cell) count, ratio of primitive cells and karyotype between the two groups. These results suggest that CD33 expression is higher in patients with high WBCs, a high proportion of primitive cells and a normal karyotype.
Gene analysis. The initial evaluation of AML patients is important and it has two prerequisite aims. On the one hand, to characterize the disease process and provide prognostic information, we can base the analysis on factors such as prior toxic exposure, antecedent myelodysplasia, and karyotypic or molecular abnormalities. These factors may impact responsiveness to chemotherapy and the risk of relapse. On the other hand, we assessed the patients' ability to tolerate chemotherapy by evaluating comorbidities. Both factors are taken into consideration when deciding treatment. Molecular abnormalities play an important role in prognostic strati cation. We found molecular abnormalities, including mutant genes and fusion genes, by performing PCR (polymerase chain reaction) and NGS (next-generation sequencing) experiments. Samples for mutation detection were collected from de novo AML patients.
To determine the gene mutation or fusion gene in relation to the level of CD33 expression, we analyzed 86 patients in this retrospective study. In 62 patients (72.1%), molecular abnormalities were detected, which may greatly contribute to the development of AML. Some genes with a high frequency among these gene abnormalities, such as FLT3, NPM1, DNMT3A, IDH1, and CEBPA, are shown in    Acute myeloid leukemia (AML) is a heterogeneous hematological malignancy characterized by the proliferation of leukemia-initiating cells, and the majority of cases have a poor prognosis. It is the most common form of acute leukemia among adults, with the largest number of deaths due to leukemia each year. The median age at diagnosis is 67 years, and 54% of patients are diagnosed when they are 65 years or older (approximately one-third are diagnosed as ≥ 75 years) [13].
Considerable progress has been made in anticancer treatments and breakthroughs in immunotherapies, but conventional AML chemotherapy, including induction and consolidation therapy, has not changed signi cantly. This highlights the need to develop new ways to improve prognosis, especially refractory and relapsed AML (R-R AML) [17][18].
Allogeneic hematopoietic stem cell transplantation (allo-HSCT) may remain the sole curative option. Therefore, immunotherapies, such as ADC, CAR-T and GO, bring promise and challenge the traditional regimens given to many R-R AML patients [9,11,19].
Although immunotherapy has been widely accepted as an effective strategy in the eld of hematological malignancies, the lack of high speci city of target antigens and the heterogeneity of AML has led to slow progress in treating AML with similar strategies. Therefore, the challenge for us is to determine an antigen expressed speci cally on the surface of leukemia cells in AML [4].
In this study, the results showed that the expression level of CD33, an independent risk factor affecting prognosis, was related to the OS of patients. At diagnosis, AML patients with a higher count of white blood cells, a higher percentage of primitive cells, and normal karyotypes had higher levels of CD33 expression. Patients with FLT3 and NPM1 mutations had higher expression of CD33. These positive results suggested that patients would most likely bene t from CD33 targeted therapy when de novo AML patients had the abovementioned clinical characteristics. In our opinion, taking into account all the available data can help improve the prognosis and survival of AML patients. Limitations of this study include the sample size and grouping the patients according to more detailed characteristics would be bene cial for analysis.
Of course, more research is required, preferably prospective clinical data, to optimize the regimen of targeting CD33 for AML. More clinical features or other factors need to be identi ed to help us choose CD33 as a targeted therapy for AML or in combination with chemotherapy.
Patients with FLT3 and NPM1 mutations had higher expression of CD33, which was consistent with the published literature [20]. The high expression of CD33 in NPM1-mutated patients may be related to the concomitant occurrence of NPM1 and FLT3 mutations. However, this conclusion needs to be supported by more clinical data. We can conclude that CD33 is a target for AML immunotherapy [21].
Targeting CD33 appears to be a suitable alternative in patients who lack hematologic stem cell donors.
Ongoing efforts are needed to optimize the application to enhance therapeutic effects and decrease injury, including cytotoxicity and economic losses. To investigate the possibility of immune-based therapies beyond stem cell transplantation to treat hematologic malignancies and recommend targeting CD33 therapy for more extended list treatments, future research should focus on studies with higher quality parameters, such as larger sample sizes, randomized studies and prospective studies.
Responsive biomarkers will enable us to select patients who are more likely to bene t from immune checkpoints and monoclonal-based therapies. Exploiting the true potential of immune agents in AML requires excellently designed clinical trials. Trials are ongoing and will guide further development of immune agents [17,[21][22][23][24].

Declarations
Ethics approval and consent to participate All procedures performed in studies involving human participants were in accordance with the ethical standards of the Research Ethics Committee of The First A liated Hospital of Henan University Science and Technology and with the 1964 Helsinki declaration and its later amendments. ALL written informed consent to participate in the study was obtained from AML patients for data to be analyse.

Consent for publication
All subjects have written informed consent.

Availability of data and materials
The datasets used or analysed during the current study are available from the corresponding author on reasonable request.

Statement of con ict of interest
The authors declare that they have no competing interests.
Funding None

Authors' contributions
The authors contributed equally to this work. Jingjing Liu analyzed and interpreted the patient data.
Jiayin Tong and Haiping Yang collected patients' data. Jingjing Liu and Haiping Yang drafted and revised the manuscript. All authors have read and approved the nal manuscript and con rmed the authenticity of all the raw data.The authors are accountable for all aspects of the work and ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.