骨髓增生异常综合征 (MDS) 的诊断常常具有挑战性，特别是在与血细胞减少的其他非肿瘤原因的区别方面。目前，它是基于外周血细胞减少、外周血和骨髓发育不良/原始细胞以及克隆细胞遗传学异常的存在，但MDS的诊断特征是多态性和非特异性的。我们研究了 BC 6800 Plus 分析仪（迈瑞诊断）的全血细胞计数 (CBC) 和研究参数 (RUO) 在区分 MDS 相关血细胞减少症方面的效用。使用来自健康个体的 100 个样本来确定正常情况下的研究参数值科目。进行了一项回顾性研究，包括 66 名诊断为 MDS 的患者、90 名因其他疾病（接受治疗的癌症患者、再生障碍性贫血、其他血液恶性肿瘤）导致的血细胞减少患者和 50 名大红细胞性贫血患者。 Wilcoxon 检验用于检测各组患者之间的统计差异，认为 p < 0.05 具有显着性。使用受试者工作特征 (ROC) 曲线分析评估用于区分血细胞减少症中的 MDS 的 RUO 参数的诊断性能。采用多变量逻辑回归模型来确定 MDS 的潜在预测因子。评估了模型的曲线下面积 (AUC) 和 Hosmer-Lemeshow 检验。该模型的性能在一项包括 224 名血细胞减少患者（验证组）的前瞻性研究中得到了验证。在 MDS 组中，平均细胞体积（MCV）、大红细胞百分比（MAC）、红细胞分布宽度（RDW）和与血细胞减少和正常患者相比，未成熟血小板分数（IPF）值升高，而血小板、红细胞和白细胞计数、Neu X（与中性粒细胞的细胞质复杂性有关）、Neu Y（与核酸含量有关）和Neu Z （与细胞大小相关）较低（p < 0.001）。 Neu X、Neu Y 和 Neu Z 在检测 MDS 时表现出较高的 AUC > 0.80； MAC、RDW 和 IPF AUC > 0.76。多变量模型证明Neu X和Neu Y可以用于MDS的识别，AUC 0.88。在验证组中，89.0%的MDS患者得到了很好的分类。MDS是常见的恶性疾病，预后较差，需要早期诊断以获得最佳治疗效益。 RUO 获得检测 MDS 发育不良的见解，并可用于鉴别诊断 MDS 与其他病因引起的血细胞减少症。

The diagnosis of Myelodysplastic syndromes (MDS) is frequently challenging, especially in terms of the distinction from the other non-neoplastic causes of cytopenia. Currently, it is based on the presence of peripheral blood cytopenias, peripheral blood and bone marrow dysplasia/blasts, and clonal cytogenetic abnormalities, but MDS diagnostic features are polymorphic and non-specific. We investigated the utility of complete blood count (CBC) and research parameters (RUO) from the analyzer BC 6800 Plus (Mindray Diagnostics) to discriminate MDS-related cytopenia.100 samples from healthy individuals were used to establish the values of research parameters in normal subjects. A retrospective study was conducted including 66 patients diagnosed with MDS, 90 cytopenic patients due to other diseases (cancer patients receiving therapy, aplastic anemia, other hematological malignancies) and 50 with macrocytic anemia. The Wilcoxon test was applied to detect statistical differences among the groups of patients, considering p < 0.05 significant. The diagnostic performance of the RUO parameters for discriminating MDS among cytopenias was evaluated using receiver operating characteristic (ROC) curve analysis. Amultivariable logistic regression model was performed to identify the potential predictors for having MDS. The area under curve (AUC) and the Hosmer-Lemeshow test of the model were assessed. The performance of the model was verified in a prospective study including 224 cytopenic patients (validation group).In the MDS group, the mean cell volume (MCV), percentage of macrocytic red cells (MAC), red cell distribution width (RDW) and immature platelets fraction (IPF) had increased values compared to the cytopenic and normal patients, while platelets, red and white cell counts, Neu X (related to the cytoplasmic complexity of neutrophils), Neu Y (related to nucleic acid content) and Neu Z (related to cell size) were lower (p < 0.001). Neu X, Neu Y, and Neu Z showed higher AUC for detecting MDS > 0.80; MAC, RDW and IPF AUC > 0.76. The multivariable model demonstrated that Neu X and Neu Y could be used in the recognition of MDS, AUC 0.88. In the validation group, 89.0% of the MDS patients were well classified.MDS are common malignant disorders with a poor prognosis, and early diagnosis is warranted for optimal benefit from treatment. RUO gain insights to detect dysplasia of MDS and could be used in the differential diagnosis of MDS from cytopenias of other etiologies.