改良支气管肺泡灌洗液细胞分类计数制片及染色法

An improved preparation of cell suspension and staining method for cell differential count in bronchoalveolar lavage fluid

目的：改良支气管肺泡灌洗液细胞分类制片及染色方法，并对其临床实用性进行验证。 方法：随机抽取50例门诊及住院患者肺泡灌洗液标本，分别用改良的自动化瑞氏吉姆萨染色法及传统手工HE染色法染色，由两名具有专业技术能力的检验人员分别进行显微镜下细胞分类计数，评估两种制片方法检测结果的可比性。 结果：两种方法对肺泡灌洗液中吞噬细胞、中性粒细胞、淋巴细胞和嗜酸性粒细胞计数一致性较好，相关系数（r）均〉0.95。改良法染色的细胞形态完整，细胞核和细胞浆分辨清楚，可见颗粒、空泡等细微结构；而传统法染色后的细胞固缩明显，细胞体积变小，细胞核和细胞浆分辨清晰度下降，颗粒和空泡等细胞结构辨认不清。改良法比传统法省去了滤液离心富集细胞的过程（6 min），并且使用自动化瑞氏吉姆萨染色法（20 min）代替了手工HE染色法（30 min），每份标本至少节省了15－20 min的制片和染色时间。 结论：改良的自动化瑞氏吉姆萨染色法优于手工HE染色法，更适用于支气管肺泡灌洗液细胞分类计数，尤其是批量检测。

Objective：To improve the preparation of cell suspension and staining method of cell differential count in bronchoalveolar lavage fluid （BALF）, and evaluate its clinical practicability. Methods：We randomly collected 50 BALF samples from outpatients and inpatients, and all the samples were prepared and stained by the improved automatic Wright Giemsa staining method and traditional manual HE staining method, respectively. Then, cell differential count in BALF was performed by 2 professional technicians, respectively, and the results obtained from the two methods were compared. Results：The consistency of phagocyte, neutrophil, lymphocyte and eosinophil counts in BALF between the two methods were well, and all correlation coefficients （r） were above 0.95. The morphology of the cells stained by the improved method was well presented, with clear nucleus, granules and vacuoles, while the cells stained by HE staining with pyknotic cell body, blear margin between nucleus and cytoplasma, and obscure granules and vacuoles. Compared with HE staining, the improved method was more rapid, because it omitted the procedure of centrifuging filter liquor and enriching cells （6 min）, and reduced the staining time （from 30 min for HE staining to 20 min for automatic Wright Giemsa staining）. Conclusion：The improved automatic Wright Giemsa staining method is superior to the traditional manual HE staining method for the cell differential count in BALF, and it is more suited to the detection in batches.