Influence on therapeutic outcome of platelet count at diagnosis in patients with de novo non-APL acute myeloid leukemia

Background Platelet (PLT) count at diagnosis plays an important role in cancer development and progression in solid tumors. However, it remains controversial whether PLT count at diagnosis influences therapeutic outcome in patients with non-acute promyelocytic leukemia (APL) acute myeloid leukemia (AML). Methods This study analyzed the relationship between PLT count at diagnosis and genetic mutations in a cohort of 330 newly diagnosed non-APL AML patients. The impact of PLT count on complete remission, minimal residual disease status and relapse-free survival (RFS) were evaluated after chemotherapy or allogeneic hematopoietic stem cell transplantation (allo-HSCT). Results Our studies showed that patients with DNMT3A mutations have a higher PLT count at diagnosis, while patients with CEBPA biallelic mutations or t(8;21)(q22; q22) translocation had lower PLT count at diagnosis. Furthermore, non-APL AML patients with high platelet count (> 65 × 109/L) at diagnosis had worse response to induction chemotherapy and RFS than those with low PLT count. In addition, allo-HSCT could not absolutely attenuated the negative impact of high PLT count on the survival of non-APL AML patients. Conclusion PLT count at diagnosis has a predictive value for therapeutic outcome for non-APL AML patients. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-023-11543-5.


Introduction
Acute myeloid leukemia (AML) is one of the most common adult hematopoietic malignancies with poor prognosis [1,2].70% of AML patients achieve complete morphologic remission after standard "3 + 7" induction treatment, most patients with complete remission (CR) relapse and progress to refractory leukemia after instance, high white blood cell count (WBC) is associated with more probability of early mortality and occurrence of central nervous system leukemia [10,11].Clinically, the majority of AML patients have thrombocytopenia, leukocytosis and anemia at diagnosis, and only a small number of patients have normal or high platelet (PLT) count [12,13].Studies showed that PLTs increased the resistance of colon and ovarian cancer cell lines to 5-fluorouracil and paclitaxel [14,15].Increasing evidence suggested that PLTs played a predominant role in colon and breast cancer cells metastasis to lung and brain [16,17].Thrombocytosis was considered as an adverse-risk factor in lung, gastric, colon, breast and kidney cancers [18][19][20][21].Reportedly, platelet microparticles (PMPs) in AML was important in leukemic development and contributed to chemotherapeutic resistance [22,23].A clinical trial showed that AML patients with a medium PLT count of 50-120 × 10 9 /L had longer overall survival (OS) and disease-free survival (DFS) than the other patients [24].Other studies showed that low PLT count were associated with better survival in intermediate-risk AML patients [24,25].However, the influence of high PLT count on therapeutic outcome remains obscure in AML patients.
In this study, we investigated the relationship between clinic characteristics and PLT count, analyzed the impact of PLT count on therapeutic outcome and MRD status, followed up their survival after chemotherapy or allo-HSCT in newly diagnosed non-APL AML patients.

Treatments
First, patients were treated with standard induction chemotherapy based on the "3 + 7" regimen, including idarubicin (IDA 10 mg/m 2 , days 1-3) or daunorubicin (DNR 60 mg/m 2 , days 1-3) and cytarabine (Ara-C 100 mg/m 2 , days 1-7).Bone marrow (BM) aspiration was performed after 14-21 days post-induction chemotherapy to evaluate the treatment response.Patients with CR continued to receive two cycles of consolidation chemotherapy based on high-dose cytarabine (HD-Ara-C 2 g/m 2 , twice, days 1-3).Patients without CR received the second cycle of induction chemotherapy based on high-dose Ara-C (Ara-C 2 g/m 2 plus cladribine 5 mg/m 2 , days 1-5 and G-CSF 300 ug days 0-5).Routine blood tests were performed to provide necessary supportive treatment.Tyrosine kinase inhibitors or FLT3 inhibitors were added to the induction and consolidation treatments in Philadelphia or FLT3-internal tandem duplication (ITD)-positive AML patients.After CR, all patients received two cycles of cytarabine-based consolidation chemotherapy.For adverse-risk patients, allo-HSCT was administrated after two cycles of consolidation chemotherapy, except for those without HLA-matched donors or refusing haploidentical transplantation.For favorable and

Allo-HSCT
As described previously [27], there were two alternative myeloablative conditioning regimens in our center, including busulfan (Bu 3.2 mg/kg/day, -7 to -4 days) + cyclophosphamide (Cy 60 mg/kg/day, -3 to -2 days) and Bu (3.2 mg/kg/day, -6 to -3 days) + fludarabine (Flu 30 mg/m 2 , -6 to -2 days).Graft-versus-host disease (GVHD) prophylaxis was regularly administered, such as cyclosporine A (CsA) plus methotrexate (MTX) in HLAmatched sibling donor transplant, CsA + MTX + antithymocyte globulin and/or mycophenolate in HLA-matched unrelated donor and haploidentical transplants.CsA was gradually withdrawn after 30 days post-transplantation, and donor lymphocyte infusion was applied after 60 days post-transplantation in patients without GVHD.Effective regiments were used to keep GVHD under control, such as methylprednisolone, CsA, CD25 monoclonal antibody or combined with other immunosuppressive agents for acute GVHD, corticosteroids and CsA for chronic GVHD.

Definition of clinical end points
Treatment response was assessed by routine blood tests and BM morphology according to standardization response criteria.CR was defined as < 5% BM leukemic blasts with normal maturation of all cell lineages [28].In addition, recovery of neutrophils (≥ 1500/µl) and PLTs (≥ 100,000/µl) in peripheral blood was mandatory, with no evidence of circulating blasts and/or extramedullary leukemia.Relapse was defined as the re-occurrence of 5% leukemic blasts in BM, re-appearance of circulating blasts or development of extramedullary leukemia [28].MRD was performed by multiparametric flow cytometry to detect abnormal leukemia populations with leukemia-associated immunophenotypes in total CD45 + cells in patients with CR before each cycle of consolidation therapy.MRD-was defined as the detection of < 0.1% abnormal cells, and MRD + was defined as the detection of ≥ 0.1% abnormal cells.Patients achieved CR were followed up for 2 years to calculate their relapse-free survival (RFS).RFS was measured from the date of first CR (CR1) until death, the first relapse, or the last follow-up in continuous CR.
Median values and ranges were used for continuous variables and percentages for categorical variables.Groups were compared using the Pearson chi-square test or Fisher's exact test for categorical variables and Mann-Whitney U tests for continuous variables.The discriminatory power of PLT count value to predict CR was assessed by estimating the area under the receiver operating characteristic (ROC) curve (AUC).The optimal cut-off value was determined by maximizing sensitivity and specificity and their 95% confidence intervals (CIs).Cox proportional hazards regression models were used to determine the influence of PLT count on RFS in AML patients after chemotherapy or allo-HSCT, and the results are expressed as hazard ratio (HR) with 95% CI.All statistical tests were 2-sided, and a P value of < 0.05 was considered statistically significant.

Clinical characteristics of AML patients
A cohort of 330 patients were included in the study.PLT counts ranged from 6 to 314, with a median of 42 × 10 9 /L.The distribution of PLT counts did not suggest apparent grouping (Fig. 2a).Based on CR after the first cycle of induction chemotherapy, the cut-off value of PLT count for therapeutic outcome in AML patients according to the ROC curve analysis was 65 × 10 9 /L (Fig. 2b).Therefore, the patients were divided into low PLT count group (≤ 65 × 10 9 /L, n = 216, 65.5%) and high PLT count group (> 65 × 10 9 /L, n = 114, 34.5%).The clinical characteristics of these patients were shown in Table 1.Patients with high PLT count had higher hemoglobin levels and more megakaryocytes (MKs) in the BM (P = 0.006 and 0.001).Cytogenetics and molecular abnormalities were compared between both groups.As shown in Table 2, the high PLT count group had more patients with DNMT3A mutation, and the low PLT count group had more patients with t(8;21)(q22;q22) translocation and CEBPA biallelic mutation.In addition, we found that PLT counts in patients with t(8;21)(q22;q22) were lower than in those with normal karyotype (27 × 10 9 /L vs. 68 × 10 9 /L, P = 0.000, Fig. 2c).Patients with CEBPA biallelic mutations had lower PLT count (31 × 10 9 /L, P = 0.013, Fig. 2d), and those with DNMT3A mutations had higher PLT count (105 × 10 9 /L, P = 0.000, Fig. 2e), as compared with CEBPA and DNMT3A wild-type at diagnosis (68 × 10 9 /L and 60 × 10 9 /L).
Patients without CR received the second cycle of induction chemotherapy (n = 79); 95.4% (206/216) of patients in the low PLT group achieved CR as compared to 81.6% (93/114) in the high PLT group (P = 0.000, Fig. 3a).After two cycles of induction chemotherapy, patients in the low PLT group had a higher CR rate than the high PLT group in intermediate-risk group (92.6% vs. 69.0%,P = 0.003, Fig. 3a).However, the CR rates were not significantly different between the low and high PLT groups in favorable-risk group (99.0% vs. 100%, P = 0.628, Fig. 3a) and adverse-risk group (92.1% vs. 81.6%,P = 0.098, Fig. 3a).

Discussion
AML is a hematological malignancy with significant clinical heterogeneity.Cytogenetic abnormalities and molecular mutations are critical indicators for the prognostic stratification of AML patients, which can help formulate an optimal therapy strategy [29][30][31].In addition, clinical characteristics at diagnosis also contribute to chemotherapeutic response and survival in AML patients [10].In this study, our data showed that higher  PLT count was associated with worse response to induction chemotherapy, fewer proportion of CR/MRD-and shorter RFS in AML patients.We further noted that these effects were more profound in intermediate-risk patients than favorable and adverse-risk AML patients.These findings were consistent with other investigations that lower PLT count predicted better survival in intermediate-risk group [24,32].Moreover, hyperleukocytosis defined as WBC > 100 × 10 9 /L diagnosis was demonstrated to relate with increasing early mortality in AML patients [10,33], which prognostic significance in RFS is not widely recognized [34,35].Our data showed that WBC < 20 × 10 9 /L was a beneficial factor for RFS in AML patients.It was also reported that WBC ≥ 20 × 10 9 /L was correlated with decreasing EFS in newly diagnosed cytogenetically normal AML patients [36].Some research showed that AML patients with medium PLT count at diagnosis in the range of 50-120 × 10 9 /L had longer OS and DFS than the other patients [24].Others reported that pretreatment PLT count > 130 × 10 9 /L was an unfavorable prognostic factor for chemotherapy response and prognosis in AML patients [32].Although cut-off value of PLT count was various in different studies, these clinical data demonstrated that higher PLT count harbored a negative impact on survival of AML patients.However, the relationship between PLT count and therapeutic outcome in different risk groups was unclear.About 50% AML patients are classified as intermediate-risk group based on ELN risk classification, which 4-year OS was no more than 50% [37].Therefore, it is critical to identify novel risk factors for patients in intermediate-risk group who fail to induction therapy.Our study demonstrated that PLT count was considered as a valuable indicator to predict therapeutic response and RFS of AML patients in intermediate-risk group.
Platelets have been reported to play a pivotal role in cell proliferation, metastasis, drug resistance in lung and ovarian cancers [38,39].Higher PLT count confers poor prognosis in many cancer types, including colon, lung, ovary, and stomach [19,20,40,41].A variety of substances stored and secreted by platelets had effect on proliferation of leukemic cells, such as platelet-derived growth factor, vascular endothelial growth factor, transforming growth factor-β and serotonin [42][43][44][45][46].It was shown that PMPs could be internalized by AML cells, which transferred microRNAs in relation to chemotherapy resistance from platelets to leukemic cells via PMPs internalization [22,[47][48][49].Our study showed that PLT    count were negatively associated with chemosensitivity of AML patients in intermediate-risk group, but the mechanisms need to be further explored.Studies showed that cytogenetic or molecular abnormalities had an influence on proliferation and differentiation of MKs as well as platelets production in AML patients [32,50].The thrombopoietin (TPO)/myeloproliferative leukemia virus oncogene (MPL) pathway plays a critical role in both normal and malignant hematopoiesis and megakaryopoiesis [51,52].Furthermore, TPO/MPL pathway are involved in the interaction of human leukemic stem cells (LSCs) with hematopoietic microenvironment [53,54].Upregulation of the TPO/MPL signaling pathway protects the human LSCs from chemotherapy, which results in chemoresistance and recurrence [54].It was reported that TPO/MPL signaling was up-regulated in DNMT3A mutated AML patients with high PLT count and poor prognosis [55].Our data showed that high PLT count at diagnosis was found in DNMT3A mutated AML, which was more probability detected in M5 subtype [56,57].Interestingly, it was reported that high expression of MPL on blasts in AML with t(8;21) led to severe thrombocytopenia by scavenging TPO [25,58], which was consistent with our results.
There were limited numbers of patients in our singlecenter retrospective study, it wasn't found that normal or elevated platelet counts were frequently detected in AML patents with some genetic mutations or cytogenetic abnormalities as RUNX1 and chromosome 3q abnormalities, which were reported in other studies [59,60].Otherwise, the mechanism of influence on therapeutic outcome by platelets wasn't explored, investigations was needed in our further study.
In conclusion, our study demonstrated that higher platelet count at diagnosis was related to worse therapeutic outcome and shorter RFS in AML patients, especially in intermediate-risk AML patients.Further mechanistic investigations are needed to provide novel potential targets for AML patients.

Fig. 3
Fig. 3 (a) The proportion of CR after 1-2 cycles of induction chemotherapy between low and high PLT groups in the whole cohort of AML, favorable-risk AML (FR)-AML, intermediate-risk AML (IR-AML) and adverse-risk AML (AR-AML) patients.(b) The proportion of MRD-negative CR (CR/MRD-) after induction chemotherapy between low and high PLT groups in the whole cohort of AML, FR-AML, IR-AML and AR-AML patients.(c) The proportion of CR/MRD-after the first cycle of consolidation of chemotherapy between low and high PLT groups in the whole AML and different subgroups.(d) The proportion of CR/MRD-after two cycles of consolidation of chemotherapy between low and high PLT groups in the whole cohort and different subgroups.*P < 0.05, **P < 0.01,***P < 0.001 CR: complete remission; PLT: platelet; AML: acute myeloid leukemia; FR: favorable-risk; IR: intermediate-risk; AR: adverse-risk; MRD: minimal residual disease

Fig. 4
Fig. 4 (a-d) For whole AML patients treated with chemotherapy only, the 2-year RFS was compared between low-and high-PLT groups; (a) The 2-year RFS between low-and high-PLT groups in AML; (b) In favorable-risk AML, the 2-year RFS between low-and high-PLT groups; (c) In intermediate-risk AML, the 2-year RFS between low-and high-PLT groups; (d) In adverse-risk AML, the 2-year RFS between low-and high-PLT groups in AML; (e-h) For whole AML patients treated with allo-HSCT, the 2-year RFS was compared between low-and high-PLT groups; (e) The 2-year RFS between low-and high-PLT groups in AML; (f) In favorable-risk AML, the 2-year RFS between low-and high-PLT groups; (g) In intermediate-risk AML, the 2-year RFS between low-and high-PLT groups; (h) In adverse-risk AML, the 2-year RFS between low-and high-PLT groups in AML

Table 1
Patient characteristics

Table 2
Cytogenetics and molecular abnormalities between low and high PLT groups CEBPA bi , biallelic mutations of the CEBPA gene; CEBPA smbZIP , monoallelic mutations of the CEBPA gene in C-terminal DNA-binding or basic leucine zipper region