Paul K. Wallace [Color figure can be viewed at wileyonlinelibrary.com]

This issue of Clinical Cytometry B features an article by Neil Came and the group at the Peter MacCallum Cancer Centre in Melbourne, Australia describing how to reduce the amount of sample preparation time for MRD assays while retaining conformity with the EuroFlow pre‐lysis approach. Other reports found in this issue include an analysis of erythroid maturation patterns using non‐lysed bone marrow submitted by Anna Porwit in Sweden, and two manuscripts submitted by Marc Sorigue in Barcelona on methods to distinguish chronic lymphocytic leukemia (CLL) from other leukemic lymphoproliferative disorders. Rounding out the discussion on lymphoproliferative diseases are two articles on the use of flow cytometry in the diagnostic workup of follicular lymphoma, diffuse large B cell lymphoma (DLBCL), and Burkitt's lymphoma, one submitted by Nikolaos Tsagaraki in Greece and the other by David Azoulay in Israel. Turning to image analysis, the submission from Jaesool Shim and Migyung Cho from South Korea discusses the use of three shape specific parameters to model glands in tissue and how quantitative digital analysis is beginning to integrate with the more traditional morphological approaches.

Minimal residual disease (MRD) assessment by multicolor flow cytometry is a powerful prognostic tool and becoming part of standard patient management for a number of hematologic malignancies including chronic lymphocytic leukemia (CLL), multiple myeloma (MM), B‐cell acute lymphoblastic leukemia (B‐ALL), T‐cell acute lymphoblastic leukemia (T‐ALL), and acute myeloid leukemia (AML). This has been driven in part by the long‐lasting remissions many patients on clinical trials achieve and the need for surrogate endpoints of progression free survival to speedily evaluate novel therapeutics (1-3). To achieve the sensitivities of 10^‐5 ‐ 10^‐6 required for MRD assessment, more cells than can normally be obtained from 100 μL of blood or bone marrow are required (4). Laboratories have developed several methods to obtain the 5‐10 x 10⁶ cells per tube required for MRD assays but a standardized approach is needed (5). One method, the bulk lyse approach, described by the EuroFlow Consortium has become the de facto standard (6), but in comparison to routine methods it is labor intensive. Neil Came and colleagues at the Peter MacCallum Cancer Center and the Victorian Comprehensive Cancer Center in Victoria, Australia report in this issue on their validation of a modified EuroFlow pre‐lysis cell preparation method to more rapidly obtain sufficient WBC's for high sensitivity MRD assays. In their manuscript entitled “Validation of a Modified Pre‐Lysis Sample Preparation Technique for Flow Cytometric Minimal Residual Disease Assessment in Multiple Myeloma, Chronic Lymphocytic Leukemia, and B‐Non Hodgkin Lymphoma,” the authors described a method which eliminates several of the EuroFlow wash steps. This reduces the sample processing time by over an hour. Looking at MM, CLL, and B‐NHL they compare in parallel their modified pre‐lysis technique with their standard post lysis method. Pre‐lysis achieved at least a 10^‐5 sensitivity in 69% MM, 86% CLL and 82% of B‐NHL samples while maintaining stable leukocyte composition. They also report there was no significant difference in the MFI expression of most antigens using their bulk lysis approach except for a mild reduction in CD138, CD43, and CD45 and a gain in CD200 intensity. The authors conclude their method represents an achievable alternative for laboratories seeking to adopt a more cost‐effective flow cytometric approach to MRD assessment.

Myelodysplastic Syndrome (MDS) is usually diagnosed by morphology but multicolor flow cytometry has been well established as a helpful tool. Most recently recommendations have been put forth stating that an abnormal flow cytometric phenotype should be considered as a criterion in the diagnosis of patients with suspected MDS (7, 8). The use of flow cytometry for the diagnosis of MDS is currently limited to expert laboratories because of the difficulty analyzing and quantifying the many phenotypic abnormalities associated with this heterogenous disease (8). Anna Porwit and colleagues from Skåne and Lund University in Sweden and University of Saskatchewan in Canada describe an interesting new approach to examine erythroid maturation in patients with myelodysplastic syndrome. In their article entitled “Analysis of erythroid maturation in the non‐lysed bone marrow with help of radar‐plots facilitates detection of flow cytometric aberrations in myelodysplastic syndromes” they evaluate erythropoiesis using a no‐lyse approach with mAbs to CD36, CD71, CD105, CD117, CD13, and CD45 in a panel that included using DRAQ5 to gate on nucleated RBCs and WBCs. The authors conclude the proportion of erythropoiesis in their no‐lysis flow tube showed better agreement with morphology than the lysed flow tube and that the lysing process appeared to mostly affect the more mature erythroblasts. Using this method, the authors were able to accurately identify 85% of the MDS patients included in this study. Their success in identifying MDS patients can be attributed in part to the incorporation of Kaluza radar plots into their analysis of normal and aberrant erythropoiesis patterns. Because of the difficulty examining the numerous phenotypes associated with this heterogenous disease investigators have often used a variety of unsupervised clustering algorithms including SPADE, viSNE, and FlowSOM to enhance the diagnostic utility of flow cytometry for MDS (9-11). Members of this group previously used Kaluza radar plots to reproducibility display the differentiation and maturation of hematopoietic cells in normal bone marrow (12). Here they convincingly demonstrate how radar plots can facilitate the evaluation of erythropoietic maturation patterns.

Although CLL is usually easily diagnosed by its flow cytometric profile based on its expression of CD19, CD5, CD23, and under expression/loss of other B cell antigens, a few patients with leukemic lymphoproliferative disorders (LPD) show a combination of findings that are hard to classify. The CLLflow score alone or in combination with CD43 has been used to accurately distinguish between CLL and non‐CLL LPD (13, 14). Among the additional markers proposed to complement the CLLflow score only CD200 has proven useful (15, 16). In the one of the two reports found in this issue by Marc Sorigue at the Universitat Autònoma de Barcelona the role of CD200 in CLL, mantle cell lymphoma, and other LPD's is researched. In their paper, entitled the “Positive Predictive Value of CD200 Positivity in The Differential Diagnosis of Chronic Lymphocytic Leukemia” the available literature on CD200 and lymphoproliferative disorders are reviewed. They selected 27 manuscripts which included a total of 5,764 patients. CD200 was highly informative in the differential diagnosis between CLL and mantle cell lymphoma; 95% of CLL patients were found to be CD200 positive while only 8% of mantle cell patient were CD200 positive. This agrees with other studies of CLL which have generally shown positivity rates for CD200 above 97% and almost universal negativity in mantle cell lymphoma (17, 18). In other lymphoproliferative disorders, however they found CD200 to be suboptimal in the differential diagnosis. To further evaluate how other LPD's could be separated from CLL and mantle cell lymphoma, Marc Sorigue and colleagues in their second manuscript examine whether fibromodulin (FMOD) could assist in separating borderline LPDs from CLL. FMOD is a small leucine‐rich proteoglycan that modulates a variety of biological processes including angiogenesis, regulation of TGF‐β activity, inflammation, apoptosis, and metastatic phenotypes (19). FMOD is overexpressed by CLL lymphocytes but not by other cells in the blood compartment and, importantly for this study, not by other LPDs (20). In their manuscript entitled “FMOD Expression in Whole Blood Aids in Distinguishing between Chronic Lymphocytic Leukemia and Other Leukemic Lymphoproliferative Disorders. A Pilot Study” Sorigue et. al. determined FMOD levels by standard RT‐PCR in whole‐blood samples from three flow cytometrically‐defined groups (i) CLL, (ii) borderline LPD, and non‐CLL LPD. FMOD easily distinguished between CLL and non‐CLL LPD with perfect sensitivity and specificity. In 25/28 (89%) of the borderline LPD patients, FMOD expression levels fell between those for CLL and non‐CLL. The authors call for more studies but do conclude that their whole blood assay for FMOD can distinguish between CLL and non‐CLL patients and that their intermediate results for borderline LPD patients support their placement into phenotypic continuum between these disorders.

Single‐hit and double‐hit lymphomas with an immunoglobulin – MYC or other translocation partner with MYC have a poorer prognosis than standard DLBCL. In this issue, Nikolaos J. Tsagarakis and colleagues at the Athens Regional General Hospital Georgios Gennimatas investigate the immunophenotype of MYC rearranged lymphomas with the objective of using flow cytometry to guide the selection of probes for cytogenetic analysis and diagnosis. In their manuscript entitled “Contribution of Immunophenotype to The Investigation and Differential Diagnosis of Burkitt Lymphoma, Double‐Hit High‐Grade B‐Cell Lymphoma, and Single‐Hit MYC‐Rearranged Diffuse Large B‐cell Lymphoma” the authors review aggressive B‐cell lymphomas associated with MYC rearrangements seen at their center over the past 10 years. In agreement with others (21, 22) they found that a high percentage of lymphoma cells with a MYC rearrangement were positive for CD38. There are, however, well known issues establishing cut‐off points between a high and low percentage of cells expressing a marker compounded by significant overlap among CD38 expression levels in rearranged cases, high copy number cases, and wild type cases (23). To address this issue Tsagarakis et. al, further demonstrate that a high percentage of cases expressing CD10 and Ki67 in combination with a low bcl‐2 percentage was more characteristic of Burkitt's lymphoma and therefore suggest FISH could be safely restricted to MYC and IG. Deviation from this phenotype should raise suspicion of double‐hit lymphomas and therefore FISH probes for MYC, BCL2, BCL6, IGH, IGK, and IGL would be indicated.

Next David Azoulay and collaborators from Israel, investigate the use of DNA cell cycle analysis in differentiating among CD10+ follicular lymphoma, DLBCL, and Burkitt's lymphoma in their article entitled, “Flow Cytometry Aneuploidy and Cell Cycle Indexing as a Possible Tool for Differentiating Between CD10+ Diffused‐Large‐B‐cell‐Lymphoma and Follicular‐Lymphoma”. They found a higher rate of DNA aneuploidy and a higher DNA Index among CD10+ DLBCL as compared to follicular and Burkitt lymphomas. In contrast, the combined S phase G2M proliferative fraction was higher in Burkitt lymphoma compared to CD10+ DLBCL and follicular lymphoma and the proliferative fraction in DLBCL was significantly higher than in follicular lymphoma. Using a ROC analysis, an optimal proliferative fraction ≥ 8% was determined which differentiated between DLBCL and follicular lymphoma. The authors felt these data could be useful in differentiating between the two when an adequate biopsy was not available. Other investigators have used measures of DNA ploidy and proliferation to separate more aggressive from indolent disease. DNA ploidy has been used prognostically in patients with acute leukemia (24-27) and most recently Ki67 and BrdU incorporation into proliferating plasma cell disorders was found to correlate with event free survival in multiple myeloma (28).

Rounding out this issue is an interesting exploration by Lee et. al entitled “Multivariate Discriminant Analysis for Branching Classification of Colonic Tubular Adenoma Glands”. They were interested in the quantitative interpretation of morphological findings using computerized digital pathology. They explore the diagnostic utility of three shape‐specific parameters for branching classification of glands; the chord intersection ratio, convexity ratio, and maximum concave area ratio. This is a developing field that has seen an increasing number of geometric and spatial imaging methods offering new approaches for pattern‐based analysis (29-31). This and other similar approaches promises to integrate traditional pathology with the quantitative digital analysis and interpretation (29).

保罗·K·华莱士[颜色图可在wileyonlinelibrary.com上查看]

本期《临床细胞计数法B》刊登了Neil Came和澳大利亚墨尔本Peter MacCallum癌症中心的研究小组的一篇文章，描述了如何减少MRD分析的样品制备时间，同时又保持与EuroFlow预裂解方法的一致性。在本期中发现的其他报告包括瑞典的Anna Porwit提交的使用未溶解的骨髓对红系成熟模式的分析以及巴塞罗那的Marc Sorigue提交的关于区分慢性淋巴细胞性白血病（CLL）与其他白血病性淋巴细胞增生性疾病的方法的两份手稿。在关于淋巴增生性疾病的讨论中，有关流式细胞仪在滤泡性淋巴瘤，弥漫性大B细胞淋巴瘤（DLBCL）和伯基特氏淋巴瘤的诊断性检查中的应用，有两篇文章是完整的，一个由希腊的Nikolaos Tsagaraki提交，另一个由以色列的David Azoulay提交。关于图像分析，来自韩国的Jaesool Shim和Migyung Cho的论文讨论了使用三个特定形状的参数来对组织中的腺体进行建模，以及定量数字分析如何开始与更传统的形态学方法整合。

通过多色流式细胞术进行最小残留疾病（MRD）评估是一种强大的预后工具，并且已成为许多血液系统恶性肿瘤标准患者治疗的一部分，包括慢性淋巴细胞性白血病（CLL），多发性骨髓瘤（MM），B细胞急性淋巴细胞性白血病（ B‐ALL），T细胞急性淋巴细胞白血病（T‐ALL）和急性髓细胞性白血病（AML）。之所以如此，部分原因是许多患者在临床试验中获得了长期缓解，并且需要替代无进展生存期的终点来快速评估新疗法（1-3）。为了实现10的灵敏度^-5 - 10 ^-6为MRD评估所需，需要比通常可以为100μL血液或骨髓获得的多种细胞（4）。实验室已经开发出几种方法来获得MRD分析所需的每管5-10 x 10 ^6个细胞，但仍需要标准化方法（5）。EuroFlow联合会描述的一种方法，即大体积裂解法已成为事实上的标准（6），但与常规方法相比，它是劳动密集型的。澳大利亚维多利亚州Peter MacCallum癌症中心和维多利亚州综合癌症中心的Neil Came及其同事在本期报告中报告了他们对改良的EuroFlow裂解前细胞制备方法的验证，以更快速地获得足够的WBC用于高灵敏度MRD分析。在他们的手稿中“一种改良的裂解前样品制备技术对多发性骨髓瘤，慢性淋巴细胞性白血病和B-非霍奇金淋巴瘤的流式细胞最小残留疾病评估的验证，”作者描述了一种方法，该方法消除了一些EuroFlow洗涤步骤。这样可以将样品处理时间减少一个多小时。从MM，CLL和B-NHL来看，他们将改良的裂解前技术与标准裂解后方法平行进行比较。预裂解至少达到10 ^‐5在69％的MM，86％的CLL和82％的B-NHL样品中具有高灵敏度，同时保持稳定的白细胞组成。他们还报告说，除了CD138，CD43和CD45的轻度降低以及CD200强度的增加外，使用它们的整体裂解方法，大多数抗原的MFI表达没有显着差异。作者得出结论，对于寻求采用更具成本效益的流式细胞仪方法进行MRD评估的实验室来说，他们的方法代表了一种可行的替代方法。

骨髓增生异常综合症（MDS）通常通过形态学诊断，但多色流式细胞术已被确立为一种有用的工具。最近提出的建议指出，应将异常的流式细胞仪表型作为诊断可疑MDS患者的标准（7、8）。由于难以分析和量化与这种异质性疾病相关的许多表型异常，目前仅限于专家实验室使用流式细胞术诊断MDS（8）。来自瑞典斯科讷隆德大学和加拿大萨斯喀彻温大学的Anna Porwit及其同事描述了一种检查骨髓增生异常综合征患者红系成熟的有趣新方法。在他们的文章“借助雷达图分析未溶解的骨髓中的类红细胞成熟有助于检测骨髓增生异常综合症中的流式细胞仪畸变”他们使用一种单克隆抗体对CD36，CD71，CD105，CD117，CD13和CD45的单克隆抗体，采用无酶方法评估红细胞生成，其中包括使用DRAQ5选通成核的RBC和WBC。作者得出的结论是，非裂解流管中红细胞生成的比例与溶质流管相比，在形态学上具有更好的一致性，并且裂解过程似乎主要影响较成熟的成红细胞。使用这种方法，作者能够准确地识别本研究中包括的MDS患者的85％。他们在识别MDS患者方面的成功部分归因于将Kaluza雷达图纳入了他们对正常和异常红细胞生成模式的分析中。9-11）。该小组的成员以前曾使用Kaluza雷达图来重现性，显示了正常骨髓中造血细胞的分化和成熟（12）。在这里，他们令人信服地证明了雷达图如何促进评估促红细胞生成的成熟模式。

尽管通常基于其CD19，CD5，CD23的表达以及在其他B细胞抗原的表达/缺失下，通过流式细胞仪分析可以很容易地诊断CLL，但是一些患有白血病淋巴增生性疾病（LPD）的患者表现出以下综合发现：很难分类。单独使用CLLflow评分或结合CD43使用CLLflow评分可以准确地区分CLL和非CLL LPD（13，14）。在提出的补充CLLflow评分的其他标记中，只有CD200被证明是有用的（15、16）。在巴塞罗那自治大学的马克·索里格（Marc Sorigue）在本期中发现的两份报告之一中，研究了CD200在CLL，套细胞淋巴瘤和其他LPD中的作用。在他们的论文中，“ CD200阳性在慢性淋巴细胞白血病鉴别诊断中的阳性预测价值”综述了有关CD200和淋巴增生性疾病的现有文献。他们选择了27篇手稿，其中包括5764名患者。CD200对CLL和套细胞淋巴瘤的鉴别诊断具有很高的参考价值。发现95％的CLL患者为CD200阳性，而只有8％的套细胞患者为CD200阳性。这与CLL的其他研究一致，这些研究通常显示CD200的阳性率超过97％，并且在套细胞淋巴瘤中几乎普遍呈阴性（17，18）。然而，在其他淋巴增生性疾病中，他们发现CD200在鉴别诊断中次优。为了进一步评估如何将其他LPD从CLL和套细胞淋巴瘤中分离出来，Marc Sorigue及其同事在第二份手稿中研究了纤维调节蛋白（FMOD）是否可以帮助从CLL中分离临界LPD。FMOD是一种富含亮氨酸的小蛋白聚糖，可调节多种生物学过程，包括血管生成，调节TGF-β活性，炎症，凋亡和转移表型（19）。FMOD在CLL淋巴细胞中过表达，但在血腔中的其他细胞中过表达，对于本研究重要的是，其他LPD则不过表达FMOD（20）。在他们的手稿中“ FMOD在全血艾滋病中区分慢性淋巴细胞白血病和其他白血病淋巴细胞增生性疾病的表达。试点研究” Sorigue等。等 通过标准RT-PCR在三个流式细胞仪定义的组的全血样品中测定FMOD水平（i）CLL，（ii）边缘LPD和非CLL LPD。FMOD可以轻松区分CLL和非CLL LPD，具有完美的灵敏度和特异性。在25/28（89％）的临界LPD患者中，FMOD表达水平介于CLL和非CLL之间。作者呼吁进行更多研究，但得出的结论是，他们对FMOD的全血检测可以区分CLL和非CLL患者，临界LPD患者的中间结果支持他们进入这些疾病之间的表型连续体。

免疫球蛋白– MYC或具有MYC的其他易位伴侣的单发和双发淋巴瘤的预后较标准DLBCL差。在本期杂志中，雅典地区综合医院Georgios Gennimatas的Nikolaos J. Tsagarakis及其同事研究了MYC重排淋巴瘤的免疫表型，目的是使用流式细胞术指导细胞遗传学分析和诊断的探针选择。在他们的手稿“免疫表型对Burkitt淋巴瘤，双命中高级B细胞淋巴瘤和单命MYC重排弥漫性大B细胞淋巴瘤的研究和鉴别诊断中的贡献”作者回顾了过去10年中在其中心发现的与MYC重排相关的侵袭性B细胞淋巴瘤。他们与其他人一致（21、22）发现，MYC重排的淋巴瘤细胞中CD38呈阳性。有，但是，众所周知的问题，建立切断高和显著的重叠在重排的情况下，高拷贝数的情况下，与野生型病例CD38表达水平之间的复合表达标记细胞的比例很低之间的点（23）。为了解决这个问题，Tsagarakis等。等人进一步证明，高表达CD10和Ki67的病例结合低bcl-2的百分比是伯基特淋巴瘤的更多特征，因此表明FISH可以安全地局限于MYC和IG。与这种表型的偏离应引起对双重打击淋巴瘤的怀疑，因此将使用针对MYC，BCL2，BCL6，IGH，IGK和IGL的FISH探针。

接下来的David Azoulay和以色列的合作者在题为“流式细胞术非整倍性和细胞周期指数作为区分CD10 +扩散的可能工具”的文章中研究了DNA细胞周期分析在区分CD10 +滤泡性淋巴瘤，DLBCL和伯基特淋巴瘤中的用途。 “大B细胞淋巴瘤和滤泡性淋巴瘤”。他们发现，与滤泡性和伯基特淋巴瘤相比，CD10 + DLBCL的DNA非整倍性率更高，DNA指数更高。相反，与CD10 + DLBCL和滤泡性淋巴瘤相比，伯基特淋巴瘤的联合S期G2M增殖分数更高，而DLBCL的增生分数显着高于滤泡性淋巴瘤。使用ROC分析，确定了≥8％的最佳增殖分数，可区分DLBCL和滤泡性淋巴瘤。作者认为，当没有足够的活组织检查时，这些数据可能有助于区分两者。其他研究人员已经使用了DNA倍性和增殖的措施，以将更具侵略性的疾病与惰性疾病区分开。DNA倍体已被用于急性白血病患者的预后[ 24-27），最近发现Ki67和BrdU掺入到增殖性浆细胞疾病中与多发性骨髓瘤的无事件生存相关（28）。

解决这个问题是Lee等人的有趣研究。等人，题为“结肠镜管腺瘤腺体分支分类的多元判别分析”。他们对使用计算机数字病理学定量解释形态学发现感兴趣。他们探索了三种形状特定参数对腺体分支分类的诊断实用性。弦交比，凸率和最大凹率。这是一个发展中的领域，已经看到越来越多的几何和空间成像方法为基于模式的分析提供了新的方法（29-31）。这种方法和其他类似方法有望将传统病理学与定量数字分析和解释相结合（29）。