International Myeloma Working Group consensus criteria for response and minimal residual disease assessment in multiple myeloma 国际骨髓瘤工作组多发性骨髓瘤反应和最小残留病评估共识标准

Treatment of multiple myeloma has substantially changed over the past decade with the introduction of several classes of new effective drugs that have greatly improved the rates and depth of response. Response criteria in multiple myeloma were developed to use serum and urine assessment of monoclonal proteins and bone marrow assessment  (which is relatively insensitive).  Given the high rates of complete response  seen in patients with multiple myeloma with new treatment approaches, new response categories need to bedefined that can identify responses that are deeper than those conventionally defined as complete response. Recent attempts have focused on the identification of residual tumour cells in the bone marrow using flow cytometry or gene sequencing. Furthermore, sensitive imaging techniques can be used to detect the presence of residual disease outside of the bone  marrow.  Combining  these  new  methods,  the  International  Myeloma  Working  Group  has  defined  new response categories of minimal residual disease negativity, with or without imaging-based absence of extramedullary disease, to allow uniform reporting within and outside clinical trials. In this Review, we clarify several aspects of disease response assessment, along with endpoints for clinical trials, and highlight future directions for disease response assessments. 过去十年来，多发性骨髓瘤的治疗发生了重大变化，几类新药的问世大大提高了反应率和反应深度。多发性骨髓瘤的应答标准是通过血清和尿液中单克隆蛋白的评估以及骨髓评估（相对不敏感）制定的。 鉴于采用新治疗方法的多发性骨髓瘤患者完全应答率很高，因此需要确定新的应答类别，以识别比传统完全应答更深入的应答。最近的尝试主要集中在利用流式细胞术或基因测序来识别骨髓中残留的肿瘤细胞。此外，敏感的成像技术也可用于检测骨髓外是否存在残留疾病。 结合这些新方法，国际骨髓瘤工作组确定了新的反应类别，即最小残留病阴性，以及是否存在基于影像学的髓外疾病，以便在临床试验内外进行统一报告。在本综述中，我们将阐明疾病反应评估的几个方面以及临床试验的终点，并强调疾病反应评估的未来方向。

 Introduction 导言

The treatment landscape for multiple myeloma has been radically  transformed  during  the  past  decade  by  the introduction   of   several   new   drugs   with    different mechanisms  of  action,  which  has  led  to  improved survival for patients with multiple myeloma. Progress has been made in other areas, including an improved understanding of disease biology, enhanced diagnostic criteria,  availability  of  sensitive  and  specific  tools  for disease prognostication, increasingly effective treatment strategies,  and  enhanced  supportive care. The  most recent iteration of the response criteria was developed in   2006  by   the   International   Myeloma   Working Group   (appendix).   Response   evaluation  in  multiple myeloma has traditionally been based on the assessment of serum and urine monoclonal protein concentrations via protein electrophoresis or immunofixation, or both, as  a  surrogate  for  tumour  burden,  allowing  for  the detection of trace amounts of paraprotein.11 The response criteria for multiple myeloma have evolved considerably since then with the substitution of monoclonal protein concentrations for synthetic rates and the use of different cutoffs for monoclonal protein concentrations, as well as inclusion  of serum  free light  chain  (sFLC)  values  for the assessment of oligo-secretory myeloma. Traditional quantitation of bone marrow plasma cells was performed on trephine biopsies (with a combination of haemotoxylin and eosin  stains and immunohistochemistry) or bone marrow   aspirates   (with   or   without    clot   section). The importance of bone marrow plasma-cell quantitation for accurate response assessment (even inpatients with negative  serum  and  urine  immunofixation)  has  been confirmed. The   original   definition   of   a   complete response  only  required  bone  marrow  with  less  than 5%  plasma  cells,  irrespective  of their  clonal  nature. The definition was further refined to stringent complete response,  by  the  addition  of  the   sFLC   assay  plus immunohistochemical    clonal    assessment    on    the trephine  biopsy. Additional  clarifications,  especially with respect to the use of sFLC, were introduced during the International Myeloma Workshop in 2011. 过去十年间，多发性骨髓瘤的治疗格局发生了根本性的变化，几种具有不同作用机制的新药问世，改善了多发性骨髓瘤患者的生存状况。其他领域也取得了进展，包括对疾病生物学的进一步了解、诊断标准的提高、用于疾病预后判断的灵敏而具体的工具的出现、日益有效的治疗策略以及支持性护理的加强。国际骨髓瘤工作组于 2006 年制定了最新的反应标准（附录）。 多发性骨髓瘤的反应评估传统上是通过蛋白电泳或免疫吸附或两者兼用来评估血清和尿液中的单克隆蛋白浓度，以此作为肿瘤负荷的替代物，并可检测到微量的副蛋白。此后，多发性骨髓瘤的反应标准有了很大的发展，用单克隆蛋白浓度代替合成率，用不同的切点代替单克隆蛋白浓度，并将血清游离轻链（sFLC）值纳入寡分泌骨髓瘤的评估中。骨髓浆细胞的传统定量方法是对穿刺活检组织（结合血沉和伊红染色以及免疫组织化学）或骨髓抽吸物（有或无血块切片）进行定量。骨髓浆细胞定量对于准确评估反应（即使是血清和尿液免疫反应阴性的患者）的重要性已得到证实。最初的完全反应定义只要求骨髓中浆细胞少于 5%，而不论其克隆性质如何。通过添加 sFLC 检测和免疫组化克隆评估，对骨髓穿刺活检的定义进一步调整为严格的完全反应。在2011年的国际骨髓瘤研讨会上，还介绍了更多的克隆评估方法，尤其是关于sFLC的使用。

The consensus criteria were uniformly incorporated into  clinical trials,  allowing  improved  comparison  of different  drugs,  drug  combinations,  and  treatment strategies, and the revisions over the years have allowed them   to   remain   applicable   despite   advances   in treatment. With older therapies, including autologous stem-cell  transplantation   (ASCT),  less  than  half  of patients  achieve  a  complete  response. With  the introduction of more effective multidrug combinations in the past 15 years, especially when used with ASCT, post-transplant  consolidation,  and  prolonged  main- tenance therapy, nearly all patients achieve a treatment response,  with   more  than   50%   of  these   patients reaching  a  complete  response  in  some  studies. Frustratingly, most patients relapse despite achieving such  deep  responses,  reflecting  a  persistent  disease that cannot be detected with the recommended disease evaluation techniques. Consequently, new methods are urgently  required to  detect  and  quantify the level  of minimal residual disease beyond the detection of the present clinical response criteria, and the definition of disease response needs to be revised for it to evolve with the changing treatment framework. In this Review, we  report  the  new  International  Myeloma  Working Group   consensus    criteria   for   redefining   disease response with a particular emphasis on the definitions and methods to assess minimal residual disease. 共识标准被统一纳入临床试验，从而改进了不同药物、药物组合和治疗策略之间的比较。采用自体干细胞移植（ASCT）等旧疗法，只有不到一半的患者能获得完全缓解。过去15年中，随着更有效的多药联合疗法的引入，特别是与自体干细胞移植、移植后巩固治疗和长期维持治疗一起使用时，几乎所有患者都获得了治疗应答，在一些研究中，这些患者中有50%以上获得了完全应答。令人沮丧的是，大多数患者尽管获得了如此深入的应答，但病情仍会复发，这反映了一种顽固性疾病，而推荐的疾病评估技术无法检测到这种疾病。因此，迫切需要新的方法来检测和量化目前临床反应标准检测不到的最小残留病灶水平，同时需要对疾病反应的定义进行修订，使其与不断变化的治疗框架相适应。在这篇综述中，我们报告了国际骨髓瘤工作组关于重新确定疾病反应的新共识标准，并特别强调了评估极小残留病的定义和方法。

Depth of response and long-term outcome 反应深度和长期结果

The   association   between   depth   of   response   and long-term outcomes is a hotly debated topic in multiple myeloma. This debate has been particularly contentious for   complete   response,   which   has   been   generally considered as the deepest response level and a surrogate for  improved  outcome  after  any  given  treatment . The   relationship   between   complete   response   and progression-free  survival,  or  time-to-progression,  has been  more  consistent  than  the  relationship  between complete response and overall survival. This association is  frequently  seen  in  cancer  therapy  and  is  probably caused   by   multiple   factors,   including   interactions between disease biology, different treatment strategies after reaching complete response, and the true depth of response  beyond  the  conventional  (and  low-sensitive) approaches defining complete response after different therapies.  Nevertheless,  a  meta-analysis  reported  a significant  correlation  between  the  achievement  of  a complete  response  and  improved  overall  survival  in eight out often studies. Several studies using sensitive new  techniques  have  been  able  to  demonstrate  the presence of minimal residual disease that is not detected by current complete response (and stringent complete response) evaluation methods in a large proportion of patients. The  level  of  minimal   residual  disease, undetected by conventional methods, is probably one of the  most  important  features  contributing  to  the  link between the depth of response and long-term outcomes. Independent  of the  method  used  to  define  minimal residual    disease     (cell-based,    molecular-based,    or imaging-based), previous studies consistently show that among  patients  who  achieve  a  complete  response, minimal  residual   disease-positive   cases   consistently have an inferior progression-free survival than minimal residual  disease-negative  patients. Given  the  sub- stantial  proportion  of  patients  achieving  a  complete response with current therapies, response criteria need to  be  expanded  to  define  minimal  residual  disease accurately for patients with multiple myeloma. 在多发性骨髓瘤中，反应深度与长期疗效之间的关系是一个备受争议的话题。这一争论在完全应答方面尤为激烈，因为完全应答通常被认为是最深度的应答，也是任何特定治疗后疗效改善的代用指标。完全缓解与无进展生存期或进展时间之间的关系比完全缓解与总生存期之间的关系更为一致。这种关系在癌症治疗中经常出现，可能是由多种因素造成的，包括疾病生物学之间的相互作用、达到完全应答后的不同治疗策略，以及超出传统（和低敏感）方法确定不同疗法后完全应答的真正应答深度。 尽管如此，一项荟萃分析报告显示，在八项研究中，完全应答的实现与总生存率的提高之间存在显著相关性。一些使用敏感新技术的研究已经能够证明，在很大一部分患者中，存在着目前的完全应答（和严格的完全应答）评估方法无法检测到的极小残留病灶。传统方法无法检测到的极小残留病的程度，可能是导致反应深度与长期预后之间联系的最重要特征之一。无论采用哪种方法来确定极小残留病（基于细胞、基于分子或基于成像），以往的研究一致表明，在获得完全应答的患者中，极小残留病阳性病例的无进展生存期始终低于极小残留病阴性患者。鉴于目前疗法中获得完全应答的患者比例较低，因此需要扩展应答标准，以准确界定多发性骨髓瘤患者的极小残留病。

Detection of minimal residual disease in bone marrow 检测骨髓中的最小残留病灶

Bone marrow examination has been the cornerstone of disease  assessment  in  the  absence  of a  measurable monoclonal  protein  in  the  serum  or  urine,  whether this   represents   non-secretory   disease   or   complete response to therapy (ie, complete response or stringent complete  response). Increasingly,  sensitive  assays have been adopted for the evaluation of bone marrow aspirates,  including  multiparametric  flow  cytometry (MFC), allele-specific oligonucleotide (ASO)-qPCR and next-generation  sequencing  of VDJ  sequences, in  an effort  to  increase  the  sensitivity  of the  detection  of multiple  myeloma  cells. Such  methods  allow the quick examination of several hundreds of thousands to millions of bone marrow cells (or the corresponding amount   of   DNA)   per   assay   and   can   provide   a quantitative assessment of any residual tumour cells in the bone marrow. 在血清或尿液中没有可测量的单克隆蛋白的情况下，骨髓检查一直是疾病评估的基石，无论这代表的是非分泌性疾病还是对治疗的完全反应（即完全反应或严格完全反应）。为了提高检测多发性骨髓瘤细胞的灵敏度，人们越来越多地采用灵敏的检测方法来评估骨髓穿刺物，包括多参数流式细胞术（MFC）、等位基因特异性寡核苷酸（ASO）-qPCR和VDJ序列的下一代测序。通过这些方法，每次检测可快速检测数十万至数百万个骨髓细胞（或相应数量的DNA），并可对骨髓中的任何残留肿瘤细胞进行定量评估。

MFC methods for minimal residual disease detection 用于检测极小残留病的 MFC 方法

First-generation methods 第一代方法

MFC   is   now   a   key   tool   in   the   management   of haematological   malignancies,   and   improvements   in technology have increased the number of fluorochromes that can be used simultaneously and the number of cells that  can be  interrogated. This  advance  allows  a  large number of cell types, or different characteristics of the same cell type, to be studied concurrently in a fast and efficient  way.  Although   MFC-based  assessment  of bone marrow has been done in multiple myeloma for a number of years, the technology has only recently gained wide acceptance in the past decade for routine testing of patients with multiple myeloma.  MFC is now an integral   part   of   laboratory   investigations   and   the management of plasma-cell disorders, and can play an important part in the diagnosis, prognostic stratification, and  monitoring  of  response  to  therapy  via  minimal residual disease detection, the understanding of the biology of disease progression,26,48–50  the study of the role  of the  tumour  microenvironment  in  plasma  cell disorders, and the identification of potential therapeutic targets on the malignant plasma cell. Many  surface markers have been  described  for the  identification  of plasma cells and for distinguishing multiple myeloma plasma  cells  from  normal  plasma  cells.  The  most commonly used surface markers used for discriminating and categorizing normal and multiple myeloma plasma cells  include  CD138,  CD38,  CD45,  CD56,  CD19,  and cytoplasmic   κ  and  λ  immunoglobulin  light  chains. Additional   markers,   many   of  which   are   aberrantly expressed on multiple myeloma plasma cells, are also of value and include CD20, CD27, CD28, CD81, CD117, and CD200. Other markers that are being studied include CD54, CD229, CD319, and VS38c, some of which could help with plasma cell recognition inpatients undergoing therapy  with  monoclonal  antibodies  against  CD38  or CD138.   However,   in   view   of  the   heterogeneity   of expression of these markers and differences in both the number of events studied and in the analytical strategies used,  substantial  confusion  and  inconsistent  clinical interpretation  of  results  from  different   studies  has occurred. Attempts   have   been   made   to   develop consensus  guidelines  to  standardise  the  MFC-based assessment of disease in multiple myeloma and other plasma cell-related disorders. 微量荧光显微镜现在是治疗血液恶性肿瘤的重要工具，技术的进步增加了可同时使用的荧光素的数量和可检测的细胞数量。这一进步使得大量细胞类型或同一细胞类型的不同特征能够以快速、高效的方式同时进行研究。 虽然基于 MFC 的骨髓评估已在多发性骨髓瘤中应用多年，但这项技术直到最近十年才被广泛接受，用于多发性骨髓瘤患者的常规检测。 目前，骨髓造血干细胞检测已成为浆细胞疾病实验室检查和治疗中不可或缺的一部分，在诊断、预后分层、通过最小残留病检测监测治疗反应、了解疾病进展的生物学过程、26,48-50 研究肿瘤微环境在浆细胞疾病中的作用以及确定恶性浆细胞的潜在治疗靶点等方面发挥着重要作用。许多表面标记物已被描述用于识别浆细胞和区分多发性骨髓瘤浆细胞与正常浆细胞。 最常用于区分正常和多发性骨髓瘤浆细胞的表面标记包括 CD138、CD38、CD45、CD56、CD19 以及细胞质κ和λ免疫球蛋白轻链。其他标记物（其中许多在多发性骨髓瘤浆细胞上异常表达）也很有价值，包括 CD20、CD27、CD28、CD81、CD117 和 CD200。正在研究的其他标志物包括 CD54、CD229、CD319 和 VS38c，其中一些标志物有助于正在接受 CD38 或 CD138 单克隆抗体治疗的患者识别浆细胞。 然而，由于这些标记物的表达具有异质性，而且研究的事件数量和使用的分析策略也不尽相同，因此对不同研究结果的临床解释出现了很大的混乱和不一致。目前已尝试制定共识指南，以规范基于MFC的多发性骨髓瘤和其他浆细胞相关疾病的疾病评估。

Several studies have demonstrated the use of MFC in the detection of minimal residual disease in the bone marrow  (table  1).  In  a  study  of  flow-based  minimal residual   disease   assessment   in   multiple   myeloma, Rawstron and colleagues used a sensitive MFC assay that quantified normal and neoplastic plasma cells in the bone  marrow  of 45 patients who  received ASCT. Monoclonal plasma cells were detectable 3 months after ASCT in 19 (42%) of 45 patients, in whom the median progression-free  survival  was  20  months  compared with  35 months for those with undetectable multiple myeloma plasma cells. The sensitivity of the flow assay was highlighted by the presence of detectable plasma cells  in  nearly  a  third  of the  patients  with  negative immunofixation    results    and    patients    who    were minimal residual disease-positive had a worse outcome. San Miguel and colleagues reported almost identical results.  Subsequently,  larger  prospective  studies have reproduced  these  initial  observations.  The   Spanish Myeloma  Group  (PETHEMA/GEM)  used  four-colour flow  cytometry  to  study  minimal  residual  disease  in 295 patients newly diagnosed with multiple myeloma receiving   uniform   treatment   including   ASCT,   and showed that minimal residual disease was one of the most   important   predictors   of   outcome.  Minimal residual   disease  negativity   at   day   100   after  ASCT correlated with improved progression-free survival and overall survival and, furthermore, the effect of minimal residual disease negativity was equally relevant among patients  that  had  achieved  a  conventional  complete response. Similarly, Rawstron and colleagues evaluated the role of six-colour MFC in the assessment of minimal residual disease at various stages of therapy inpatients with  newly  diagnosed  multiple  myeloma  enrolled  on the MRC IX clinical trial. Among patients undergoing an ASCT, absence of minimal residual disease at day 100 was associated with statistically significantly improved progression-free  survival,  irrespective  of  cytogenetics or  achievement  of  a  complete  response.   Paiva  and colleagues studied  a  series  of  241  patients  enrolled in  the   Spanish  GEM2000  and  GEM2005MENOS65 studies. They identified the best independent predictors of early  relapse  after  achieving  a  complete  response were   persistent    minimal   residual   disease,   using four-colour  flow  cytometry  at  day  100  after  ASCT, presence of baseline high-risk cytogenetics, by use of fluorescence in-situ hybridisation. Early  relapse after achieving a complete response was associated with very poor survival in this group of patients, as was previously reported  by  Barlogie  and  colleagues. These  results again highlight the close association between disease biology   and   depth   of   response   after   therapy   in determining long-term outcomes, but also highlight the immediate  identification  of  patients  with  imminent relapse and poor survival (≤2 years). Of note, in all these studies,   three   six-colour   MFC   approaches   with   a sensitivity  of  one  in  10^4 myeloma  cells  were  used. The Intergroupe Francophone du Myelome reported on a phase 2 study of 31 patients treated with three induction cycles  of  lenalidomide,  bortezomib,  dexamethasone (RVD)  followed  by  cyclophosphamide  harvest,  ASCT, and  then  two  RVD  consolidation  cycles  followed  by 1 year of lenalidomide maintenance. Overall, 18 (58%) of  31  patients  achieved  a  complete  response,  with 21  (68%)  being  minimal  residual  disease-negative  as measured   by   MFC.   With   a   median   follow-up   of 39   months,   the   estimated    3-year   progression-free survival for the whole patient series was 77% and overall survival was 100%. None of the patients who achieved minimal  residual  disease  negativity  relapsed  after  a median of 39 months. 多项研究表明，MFC 可用于检测骨髓中的最小残留病（表 1）。 在一项基于血流的多发性骨髓瘤最小残留病评估研究中，Rawstron 及其同事使用了一种灵敏的 MFC 检测方法，对 45 名接受 ASCT 的患者骨髓中的正常和肿瘤性浆细胞进行定量。45 名患者中有 19 人（42%）在接受 ASCT 3 个月后检测到单克隆浆细胞，这些患者的中位无进展生存期为 20 个月，而检测不到多发性骨髓瘤浆细胞的患者的中位无进展生存期为 35 个月。近三分之一免疫检测结果为阴性的患者体内存在可检测到的浆细胞，这凸显了血流检测的敏感性，而极小残留病变阳性的患者预后较差。圣米格尔及其同事报告了几乎相同的结果。 随后，更大规模的前瞻性研究再现了这些初步观察结果。 西班牙骨髓瘤小组（PETHEMA/GEM）使用四色流式细胞术研究了295名接受包括ASCT在内的统一治疗的新确诊多发性骨髓瘤患者的极小残留病，结果显示极小残留病是预测预后的最重要因素之一。 ASCT后第100天的最小残留病阴性与无进展生存期和总生存期的改善相关，此外，最小残留病阴性的影响与获得常规完全反应的患者同样相关。同样，Rawstron 及其同事评估了六色 MFC 在新诊断的多发性骨髓瘤患者接受 MRC IX 临床试验的不同治疗阶段评估最小残留病的作用。在接受 ASCT 治疗的患者中，无论细胞遗传学如何，也无论是否获得完全反应，第 100 天时无极小残留病都与无进展生存期的显著提高有统计学关系。 Paiva 及其同事对参加西班牙 GEM2000 和 GEM2005MENOS65 研究的 241 名患者进行了一系列研究。他们发现，获得完全缓解后早期复发的最佳独立预测因素是持续性微小残留病（采用 ASCT 后第 100 天的四色荧光细胞测定法）和基线高风险细胞遗传学（采用荧光原位杂交法）。正如Barlogie及其同事之前报道的那样，该组患者在获得完全缓解后的早期复发与生存率极低有关。 这些结果再次凸显了疾病生物学特性与治疗后反应深度之间的密切联系对长期疗效的决定作用，同时也强调了对即将复发和生存期较短（≤2 年）患者的即时识别。值得注意的是，所有这些研究都使用了三种六色 MFC 方法，灵敏度为 10^4 骨髓瘤细胞中的一个。法语骨髓瘤小组（Intergroupe Francophone du Myelome）报告了一项2期研究，31名患者接受了来那度胺、硼替佐米、地塞米松（RVD）三个诱导周期的治疗，随后接受环磷酰胺收获、ASCT治疗，然后接受两个RVD巩固周期的治疗，再维持一年的来那度胺治疗。总体而言，31名患者中有18人（58%）获得了完全应答，其中21人（68%）通过MFC检测为最小残留病阴性。 中位随访时间为 39 个月，估计整个患者系列的 3 年无进展生存率为 77%，总生存率为 100%。在中位随访 39 个月后，没有一名极小残留病灶阴性患者复发。

The    advances    in    MFC    technology    that    allow interrogation of several million cells have significantly improved the sensitivity of the assay, particularly when combined  with  the  use  of eight  or  more  colours  or markers  for  increased  specificity.  Current  consensus indicates that such approaches are optimally suited for minimal residual disease testing of multiple myeloma. MFC 技术的进步可检测数百万个细胞，大大提高了检测灵敏度，尤其是在结合使用八种或更多颜色或标记物以提高特异性的情况下。 目前的共识表明，这种方法最适合用于多发性骨髓瘤的最小残留病检测。

In addition, MFC-based assessment of the post-therapy bone marrow provides important information regarding the immune-cell profile, which can provide additional prognostic information. A report58  from the PETHEMA group  showed  that  normal  plasma-cell  recovery  and normal B-cell maturation was associated with improved survival outcomes irrespective of the minimal residual disease status. 此外，基于 MFC 的治疗后骨髓评估可提供有关免疫细胞原的重要信息，从而提供额外的预后信息。PETHEMA 小组的一份报告58 显示，无论疾病的最小残留状态如何，浆细胞的正常恢复和 B 细胞的正常成熟都与生存率的提高有关。

Next-generation flow 下一代流式

Attempts to standardise and automate readouts for MFC make it a potentially attractive test for sensitive, routine detection  of  minimal  residual  disease  in  the  bone marrow compartment.  However, to have a uniform MFC-based minimal residual disease response criteria, consensus  in  the  way  minimal  residual  disease  is evaluated will be mandatory. Accordingly, a concerted effort  has  been  made  to  standardise  the  flow-based approaches and remove subjectivity by defining reagent characteristics,  defining  the  acquisition  and  plasma- cell  identification variables,  and by  introducing  novel common data analysis tools. 对 MFC 读数进行标准化和自动化的尝试使其成为一种具有潜在吸引力的检测方法，可用于灵敏、常规地检测骨髓区的极小残留病。 然而，要制定统一的基于 MFC 的极小残留病反应标准，就必须就极小残留病的评估方法达成共识。因此，通过确定试剂特性、确定采集和血浆细胞鉴定变量以及引入新型通用数据分析工具，我们正在共同努力使基于流程的方法标准化，并消除主观性。

The current EuroFlow next-generation flow method for minimal residual disease detection in multiple myeloma relies  on two  eight-colour  combinations that  combine surface antigens for the identification of phenotypically aberrant  clonal  plasma   cells  and   cytoplasmic  κ  and λ  light-chain   expression  to  confirm  their  clonality. The technique has been modified to include an initial bulk  lysis  step  to  consistently  measure  more  than 5 × 10⁶ leucocytes per tube. The EuroFlow group has also   developed   software   algorithms   for    automated identification of clonal plasma cells (ie, minimal residual disease)  in  multiple  myeloma  samples. This two-tube next-generation flow approach has now been extensively validated (ie, >1000 minimal residual disease samples). It is very robust and improves reliability, consistency, and sensitivity because of the acquisition of a greater number of cells. The eight-colour technology is widely available globally and the next-generation flow method has already been fully adopted by multiple flow laboratories. 目前用于检测多发性骨髓瘤极小残留病的 EuroFlow 下一代流式检测法采用两种八色组合，结合表面抗原识别表型异常的克隆性浆细胞，并结合胞质κ和λ轻链表达确定其克隆性。该技术经过改良，增加了一个初始批量裂解步骤，以保证每支试管能检测到超过 5 × 10⁶ 的白细胞。EuroFlow 小组还开发了用于自动识别多发性骨髓瘤样本中克隆浆细胞（即最小残留病）的软件算法。这种双管下一代流式方法现已得到广泛验证（即超过 1000 个极小残留病样本）。这种方法非常稳健，而且由于采集的细胞数量更多，因此提高了可靠性、一致性和灵敏度。八色技术已在全球广泛应用，下一代流式方法已被多个流式实验室全面采用。

The complete eight-colour method can be done using individual antibodies or made more efficient by using a lyophilised  mixture  of the  required  antibodies,  which reduces errors, time, and costs. Ongoing quality-control assessment  is  required  for  all  laboratories  reporting minimal  residual   disease  flow  results.   Use   of  the automated software package is ideal because it makes the method user independent, identifies and counts all bone marrow cell subsets in addition to plasma cells, and reinforces the adoption of standard operation procedures for accurate and automated readouts of patient samples. Although many advantages are lost when not using this software,  the   method  can  remain  satisfactory  when adequately validated with quality controls. 完整的八色方法可以使用单个抗体来完成，也可以通过使用所需抗体的冻干混合物来实现，从而减少误差、时间和成本。所有报告最小残留病流量结果的实验室都需要进行持续的质量控制评估。 使用自动化软件包是理想的选择，因为它使该方法不受用户的影响，除了浆细胞外，还能识别和计数所有骨髓细胞亚群，并加强了标准操作程序的采用，从而准确、自动地读取患者样本。虽然不使用该软件会失去许多优势，但如果通过质量控制进行充分验证，该方法仍能令人满意。

One of the most attractive features of the eight-colour method is its balance between effectiveness (ie, sensitivity plus specificity) and wide availability, because eight-colour instruments  are  commonly  used  in  many  hospitals. To  improve  efficiency  and  to  reduce  costs,  alternate single-tube ten-colour and 14-colour methods have been suggested by some centres. The single-tube approach will undergo detailed cross-validation with reference to the next-generation flow method to allow for standardisation. Documentation    of   cross-validation    with    reference next-generation flow, ongoing quality-control assessment, routine assessment of more than 5 million mononuclear cells   to   estimate   minimal   residual   disease,   and   a sensitivity of one in 10^5 cells or higher is needed to fulfil the criteria for the next-generation flow method. 八色法最吸引人的特点之一是兼顾了高效性（即灵敏度和特异性）和广泛性，因为八色仪器在许多医院都很常用。为了提高效率和降低成本，一些中心建议交替使用单管十色法和十四色法。单管方法将参照下一代流式方法进行详细的交叉验证，以实现标准化。要达到下一代流式方法的标准，需要提供与参考下一代流式方法进行交叉验证的文件、持续的质量控制评估、对超过 500 万个单核细胞进行常规评估以估计最小残留病，以及 10^5 个细胞中有一个或更高的灵敏度。

Molecular methods for minimal residual disease detection 检测极小残留病的分子方法

ASO-qPCR

Another method that has been studied extensively in the past is ASO-qPCR, and it has been compared head-to-head with MFC assays (table 2). Use of ASO-qPCR to identify clonal  multiple  myeloma  plasma-cell-specific  immuno- globulin heavy chain (IGH) gene rearrangements allows the  detection  of very  low  levels  of  multiple  myeloma plasma  cells  with  a  sensitivity  that  can  detect  one  in 10^5 cells.  Therefore,  unlike  the  early   PCR  methods that  were  qualitative  and  semi-quantitative,  ASO-qPCR provides an accurate quantification of minimal residual disease.  ASO-qPCR  involves  making  primers  comple- mentary  to  the  junctional  region  of  the  rearranged IGH genes, which are used to interrogate bone marrow samples  at  different  times  to  determine  the  response depth.  This  step  requires  availability  of  the  baseline diagnostic sample. Bakkus and colleagues examined the usefulness of using an ASO-qPCR assay at 3–6 months post-ASCT   to    detect    minimal   residual    disease    in 67  patients.  By  using  specific  thresholds  to  define  the quantitative PCR results, the authors identified patients with minimal residual disease positivity and, subsequently, short    time    to    relapse.    Lipinski    and    co-workers retrospectively analysed the tumour load in bone marrow samples from 13 patients at the time of remission after ASCT and at the time of progression using ASO-qPCR. Progression was detected earlier with this method than with serum monoclonal protein estimation, showing the higher sensitivity of the ASO-qPCR technique. Galimberti and   colleagues examined   the   prognostic   value    of PCR-based  monitoring  of minimal  residual  disease  in 20 patients after ASCT versus non-myeloablative allogeneic transplantation.  After  ASCT,  only  three  patients  (15%) achieved PCR negativity, whereas 12 (60%) were negative after allogeneic transplantation. At 2 years, 15  (75%)  of 20 minimal residual disease-negative patients were still alive compared with five  (25%) of 20 minimal residual disease-positive   cases.   In   another    study,   130   newly diagnosed  patients  with   multiple  myeloma  from  the GEM2000/GEM2005  trials  who  achieved  a  very  good partial response to induction therapy were studied using multiplex  PCR  for  IGH  D-J,  IGK  V-J,  and  κ-deleting element rearrangements, at baseline and after induction therapy. 64 (48%) of120 minimal residual disease-negative patients had an improved median progression-free survival compared with minimal residual disease-positive patients (61 months vs 36 months; p=0·001), and improved median overall survival (not reached vs 66 months; p=0·03). Puig and colleagues33 compared minimal residual disease status using ASO-qPCR versus four-colour MFC in a large series of 170 patients from different clinical trials who achieved at least a partial response after treatment. The authors found a  significant correlation in predicting minimal residual disease between the two techniques (r=0·881; p<0·001), with  minimal  residual  disease-negativity,  using   either method,  predicting  better  progression-free  and  overall survival. However, more than half the patients could not be evaluated by the molecular  approach  either  due to the inability to detect a clone, unsuccessful  sequencing, or suboptimal  ASO-qPCR  performance.  These  technical limitations are in part due to the presence of multiple somatic mutations in the immunoglobulin genes. In these cases, primers and probes that are adapted to each patient to match the somatic hypermutations are needed. This will be particularly important in bone marrow baseline samples with   relatively   low   levels   of   plasma-cell   infiltration. Production of specific primers and probes has not been done  consistently  in  the  reported   studies,  leading  to suboptimal results in the identification and sensitivity of identified targets. In view of the substantial proportion of patients with unsuccessful PCR-based minimal residual disease estimation, the same investigators examined the capacity of CD138 selection to increase the proportion of informative patients by comparing CD138-positive selected samples with paired unselected bone marrow samples. Within    the    CD138-positive    selected    group,    VDJH rearrangements  were  detected  in  all  25  cases   (100%), compared  with  the  control   samples  in  which   VDJH rearrangements were detected in  19  (76%)  of 25 cases. After sequencing, 24 (96%) of 25 cases within the CD138 selected  group had  a  PCR  target  for  minimal  residual disease detection compared with only 15 (60%) of 25 cases in the control group.  Despite minimal residual disease evaluation by ASO-qPCR being a sensitive and specific approach,  it  is  only  applicable  in  a  low  proportion of   patients   with   multiple   myeloma   and   is   more timeconsuming than MFC. ASO-qPCR 是过去被广泛研究的另一种方法，它与 MFC 检测法进行了正面比较（表 2）。使用 ASO-qPCR 鉴定克隆性多发性骨髓瘤浆细胞特异性免疫球蛋白重链（IGH）基因重排，可以检测到极低水平的多发性骨髓瘤浆细胞，其灵敏度可达 10^5 个细胞中检测到一个。 因此，与早期的定性和半定量 PCR 方法不同，ASO-qPCR 能够准确量化最小残留病灶。 ASO-qPCR 包括制作与重排 IGH 基因交界区互补的引物，在不同时间对骨髓样本进行检测，以确定反应深度。 这一步骤需要获得基线诊断样本。Bakkus及其同事研究了在接受造血干细胞移植后3-6个月使用ASO-qPCR测定检测67名患者的最小残留病的有用性。 通过使用特定阈值对定量 PCR 结果进行判定，作者确定了极小残留病阳性的患者，随后确定了复发时间较短的患者。 Lipinski 及其合作者使用 ASO-qPCR 回顾性分析了 13 名患者在 ASCT 缓解期和进展期骨髓样本中的肿瘤负荷。与血清单克隆蛋白估测相比，这种方法更早地检测到肿瘤进展，这表明 ASO-qPCR 技术具有更高的灵敏度。Galimberti 及其同事研究了基于 PCR 的最小残留病监测对 20 名接受 ASCT 与非消融性异体移植患者的预后价值。 ASCT 后，只有 3 名患者（15%）PCR 阴性，而异体移植后有 12 名患者（60%）PCR 阴性。2 年后，20 例极小残留病阴性患者中有 15 例（75%）仍然存活，而 20 例极小残留病阳性病例中只有 5 例（25%）存活。 在另一项研究中，研究人员在基线和诱导治疗后使用多重 PCR 检测 IGH D-J、IGK V-J 和 κ-缺失元素重排，对 130 名来自 GEM2000/GEM2005 试验、对诱导治疗有很好部分反应的新诊断多发性骨髓瘤患者进行了研究。与极小残留病灶阳性患者相比，120 例极小残留病灶阴性患者中有 64 例（48%）的中位无进展生存期有所改善（61 个月 vs 36 个月；P=0-001），中位总生存期也有所改善（未达到 vs 66 个月；P=0-03）。 Puig 及其同事33 对来自不同临床试验、治疗后至少获得部分应答的 170 例患者进行了大规模系列研究，比较了 ASO-qPCR 与四色 MFC 的最小残留病灶状态。作者发现，这两种技术在预测最小残留病方面存在显著相关性（r=0-881；p<0-001），无论使用哪种方法，最小残留病阴性都能预测更好的无进展生存期和总生存期。然而，由于无法检测到克隆、测序不成功或 ASO-qPCR 性能不理想，一半以上的患者无法通过分子方法进行评估。 这些技术限制的部分原因是免疫球蛋白基因中存在多个体细胞突变。在这种情况下，需要根据每位患者的情况调整引物和探针，以匹配体细胞高突变。这对于血浆细胞浸润水平相对较低的骨髓基线样本尤为重要。在已报道的研究中，特异性引物和探针的制作并不一致，导致鉴定结果和鉴定目标的灵敏度不理想。鉴于基于 PCR 的最小残留病灶评估有相当比例的患者没有成功，同一研究者通过比较 CD138 阳性的筛选样本和配对的未筛选骨髓样本，研究了 CD138 筛选是否有能力增加有信息的患者比例。在 CD138 阳性选取组中，所有 25 个病例（100%）都检测到了 VDJH 重排，而对照组样本中，25 个病例中有 19 个（76%）检测到了 VDJH 重排。测序后，CD138 挑选组的 25 个病例中有 24 个（96%）具有用于检测极小残留病的 PCR 靶点，而对照组的 25 个病例中只有 15 个（60%）具有该靶点。 尽管通过 ASO-qPCR 评估极小残留病是一种灵敏、特异的方法，但它只适用于一小部分多发性骨髓瘤患者，而且比 MFC 更耗时。

This  has  been  recently  reiterated  by   Drandi  and colleagues,70    who  compared  qPCR  and  droplet-digital PCR for minimal residual disease assessment in multiple myeloma, acute lymphocytic leukaemia, and mantle-cell lymphoma. The investigators showed that droplet-digital PCR was less applicable and more labour intensive. Drandi 及其同事70 最近重申了这一点，他们比较了 qPCR 和液滴-数字 PCR 在多发性骨髓瘤、急性淋巴细胞白血病和套细胞淋巴瘤中的最小残留病评估。研究人员的研究表明，液滴-数字 PCR 的适用性更差，劳动强度更大。

Next-generation sequencing 新一代测序

Next-generation sequencing is of considerable interest for the  detection  of  multiple  myeloma  minimal  residual disease   in  the  bone   marrow.   Most   published   data have been generated with the LymphoSIGHT platform (Sequenta Inc, San Francisco, CA, USA), which uses sets of multiple primers for the amplification and sequencing of immunoglobulin gene segments. Specifically, genomic DNA is amplified using locus-specific primers designed for  IGH-VDJH,  IGH-DJH,  or  IGK.  Once  amplified,  the immunoglobulin   gene   DNA   is   sequenced   and   the frequencies of the different clonotypes in the sample are determined. To avoid disproportionalamplification of the IGH  and  IGK  rearrangements,  the  extensive  sets  of primers need to be attuned and validated to guarantee equal    (proportional)    amplification    of    each    target rearrangement   between    the   many   rearrangements derived  from  remaining  normal  B  cells.  Patients  with detectable multiple myeloma clones (>5%) at baseline can then be studied at subsequent timepoints to determine the   presence   and   quantity   of  that   particular   clone using  sequencing  approaches.  Ladetto  and  colleagues compared  IGH  gene-based  minimal  residual  disease detection by ASO-qPCR and next-generation sequencing to  assess  whether  next-generation  sequencing  could overcome  some  of the  limitations  of ASO-qPCR,  and further increase its sensitivity and specificity. Clonotypes identified by next-generation sequencing and ASO-qPCR were either identical or more than 97% homologous in 41 (96%) of 43 cases. Both tools had a sensitivity of about one in 10^5 cells, but next-generation sequencing had the added advantage of not requiring patient-specific primers. Previous studies show that next-generation sequencing can   achieve   a   sensitivity   of   one   in   10^6 nucleated cells. Next-generation sequencing, as with other DNA sequence-based approaches, needs a baseline sample to identify tumour-specific sequences. 下一代测序技术在检测骨髓中多发性骨髓瘤极小残留病方面颇受关注。 大多数已发表的数据都是用 LymphoSIGHT 平台（Sequenta Inc，美国加利福尼亚州旧金山）生成的，该平台使用多组引物对免疫球蛋白基因片段进行扩增和测序。具体来说，基因组 DNA 扩增使用为 IGH-VDJH、IGH-DJH 或 IGK 设计的特定位点引物。 扩增后，对免疫球蛋白基因 DNA 进行测序，并确定样本中不同克隆类型的频率。为避免IGH和IGK重排的放大比例失调，需要对大量引物进行调整和验证，以确保在来自剩余正常B细胞的众多重排中，每个目标重排都能得到等量（比例）放大。 基线可检测到多发性骨髓瘤克隆（>5%）的患者可在随后的时间点进行研究，利用测序方法确定特定克隆的存在和数量。 Ladetto 及其同事比较了基于 IGH 基因的 ASO-qPCR 和新一代测序的最小残留病检测方法，以评估新一代测序是否能克服 ASO-qPCR 的某些局限性，并进一步提高其灵敏度和特异性。在 43 个病例中，通过新一代测序和 ASO-qPCR 鉴定出的克隆型有 41 个（96%）相同或同源性超过 97%。这两种工具的灵敏度都约为 10^5 个细胞中的一个，但下一代测序的另一个优势是不需要患者专用引物。之前的研究表明，新一代测序的灵敏度可达到 10^6 个有核细胞中的一个。与其他基于DNA序列的方法一样，下一代测序需要一个基线样本来确定肿瘤的特定序列。

Martinez-Lopez   and    colleagues compared    next- generation sequencing with first-generation four-colour MFC. Bone marrow samples at baseline and from the time of very good partial response or complete response were studied by next-generation sequencing to identify a tumour clonotype at baseline and then re-evaluated for the presence of the same clonotype in the subsequent sample. Transplantation-ineligible patients were studied at the end of induction therapy, whereas patients who were younger  in  age were  studied  at  3  months  after ASCT. A dominant multiple myeloma clone could be identified at baseline in 121 (91%) of 133 patients, with IGH-VDJH    rearrangements  in  84  (69%),  IGH-DJH     in 66  (55%),  and   IGK  in  58  (48%)  of  patients  with  a dominant clone. This observation  suggests that  some clones  are  missed  by  the  next-generation  sequencing approach, most probably because of somatic mutations. Of the  121  patients  with  an  identifiable  clonotype  at baseline, 110 had follow-up samples taken. Sequencing showed that  80  (73%)  remained positive for minimal residual disease, with at least one tumour cell in 10⁶ cells. Among  the   110  patients  who   achieved  a  very  good partial  response,  those  who  had  a  minimal  residual disease-negative status (more  than one tumour cell in 10⁵  cells)  had  a  better  progression-free  survival  and overall survival compared with those who were minimal residual disease-positive. Among the group of patients with   a   complete   response,   a   higher   proportion of cases  had  minimal  residual  disease-negativity  that also associated with improved progression-free survival compared  with  patients  who  were  minimal  residual disease-positive; however, overall survival did not differ significantly. Information on minimal residual disease detection  by  MFC  was  available  in  99  patients  and 41  patients  by  ASO-qPCR  analysis,  respectively,  and the  agreement  between   sequencing  and   MFC  and ASO-qPCR  was  83%  and  85%,  respectively. Among those with different results, 12 patients were negative by MFC but were positive by sequencing; the five remaining patients had the  opposite  pattern  (MFC-positive/next- generation sequencing-negative). Korde and colleagues73 also  used  next-generation  sequencing  in  43  patients with   multiple    myeloma   treated   with    carfilzomib, lenalidomide,   and   dexamethasone,   and   observed   a 12-month progression-free survival for minimal residual disease-negative   patients   of   100%   versus   79%   for minimal  residual  disease-positive  patients   (p<0·001). The IFM2009 trial randomised 700 patients to receive either  eight  cycles  of bortezomib,  lenalidomide,  and dexamethasone (VRD; arm A), or three VRD cycles plus ASCT   followed   by   two    consolidation   VRD   cycles (arm   B).72      All   patients   then   received   lenalidomide maintenance therapy for 12 months. 289 patients were evaluated by next-generation sequencing and 475 patients with   MFC   before   maintenance   and   178   by   next- generation seqencing and 310 by MFC after completion of  maintenance   therapy.   Minimal   residual   disease detection by next-generation sequencing was feasible in 266  (92%)  of  289  patients  with  a  sensitivity  of  one tumour  cell  in  10⁶  cells.  Among  those  patients  who achieved a complete response, the 3-year progression- free  survival  was  87%  for  minimal  residual  disease- negative   patients   and   42%   for   minimal   residual disease-positive patients, pre-maintenance therapy. The corresponding  numbers  were  83%  and   30%  when minimal residual disease was tested post-maintenance. Martinez-Lopez 及其同事将新一代测序与第一代四色 MFC 进行了比较。他们通过新一代测序技术研究了基线和部分反应非常好或完全反应时的骨髓样本，以确定基线时的肿瘤克隆型，然后重新评估后续样本中是否存在相同的克隆型。不符合移植条件的患者在诱导治疗结束时进行研究，而年龄较小的患者则在ASCT后3个月进行研究。133例患者中有121例（91%）在基线时可确定为多发性骨髓瘤显性克隆，其中84例（69%）有IGH-VDJH重排，66例（55%）有IGH-DJH重排，58例（48%）有IGK重排。这一观察结果表明，下一代测序方法漏掉了一些克隆，很可能是因为体细胞突变。在基线克隆型可确定的 121 名患者中，有 110 人接受了后续样本采集。测序结果显示，80 名患者（73%）的极小残留病灶仍呈阳性，10⁶个细胞中至少有一个肿瘤细胞。在获得很好部分反应的 110 名患者中，与极小残留病灶阳性患者相比，极小残留病灶阴性患者（10⁵个细胞中至少有一个肿瘤细胞）的无进展生存期和总生存期更长。在完全反应患者中，极小残留病灶阴性患者的比例较高，与极小残留病灶阳性患者相比，无进展生存期也有所改善；但总生存期没有显著差异。通过MFC和ASO-qPCR分析，分别有99名和41名患者获得了极小残留病检测信息，测序与MFC和ASO-qPCR的一致性分别为83%和85%。在结果不同的患者中，12 例患者的 MFC 分析结果为阴性，但测序结果为阳性；其余 5 例患者的情况正好相反（MFC 阳性/新一代测序阴性）。Korde及其同事73也在43例接受卡唑米、来那度胺和地塞米松治疗的多发性骨髓瘤患者中使用了新一代测序技术，并观察到最小残留病阴性患者的12个月无进展生存率为100%，而最小残留病阳性患者的12个月无进展生存率为79%（P<0-001）。 IFM2009 试验将 700 名患者随机分为两组，一组接受 8 个周期的硼替佐米、来那度胺和地塞米松治疗（VRD；A 组），另一组接受 3 个 VRD 周期加 ASCT，然后再接受 2 个 VRD 周期的巩固治疗（B 组）72。在维持治疗前，对 289 例患者进行了新一代测序评估，475 例患者进行了 MFC 评估；在完成维持治疗后，对 178 例患者进行了新一代测序评估，310 例患者进行了 MFC 评估。 289 例患者中有 266 例（92%）可以通过新一代测序检测出最小残留病，灵敏度为 10⁶ 个细胞中有一个肿瘤细胞。 在获得完全应答的患者中，最小残留病阴性患者的3年无进展生存率为87%，最小残留病阳性患者的3年无进展生存率为42%。在维持治疗后检测极小残留病时，相应的数字分别为 83% 和 30%。

A formal comparison with  MFC with next-generation sequencing  cannot  be  done  given  the  low  sensitivity (one tumour cell in 10⁵ cells) for the MFC method used in this study. 由于本研究中使用的 MFC 方法灵敏度较低（10⁵ 个细胞中有一个肿瘤细胞），因此无法将 MFC 与新一代测序进行正式比较。

 Comparison of techniques 技术比较

As  described  previously, various techniques have been studied  for  the  detection  of minimal  residual  disease. Each  of  these  techniques  (based  on  the  plasma-cell phenotype,  or  genotype,  or both) have  advantages  and disadvantages that need to be taken into consideration (table 3). The ideal minimal residual disease test should fulfil  several  relevant  characteristics:  high  applicability (useful in most patients), high sensitivity and specificity, excellent  feasibility  (result  can  be  obtained  in  most patients),   easily   accessible,   rapid   turnaround,   small sample   size   that   can   be   transported   with   relative ease,    reproducibility,    proven     clinical    value,     and cost-effectiveness. A notable disadvantage of the sequence- based approach is the requirement of a baseline sample to identify tumour-specific sequences. 如前所述，目前已研究出多种用于检测极小残留病的技术。这些技术（基于血浆细胞表型或基因型，或两者兼而有之）各有利弊，需要加以考虑（表 3）。理想的极小残留病检测应具备以下几个相关特征：适用性强（对大多数患者有用）、灵敏度和特异性高、可行性好（大多数患者都能得到结果）、容易获得、周转快、样本量少且运输相对方便、可重复性好、临床价值得到证实以及成本效益高。基于序列的方法的一个显著缺点是需要基线样本来识别肿瘤特定序列。

While no currently available tests fully satisfy all these ideal  criteria,  next-generation  sequencing  and  next- generation flow fulfil most of them and can be translated into an advanced platform that can be uniformly applied across institutions and countries. Next-generation sequencing and next-generation flow have been reported to have variable levels of sensitivity. Both methods have the  ability  to  detect  one  multiple  myeloma  cell  in 10⁵–10⁶ cells. We  strongly encourage the inclusion of both methods in prospective trials, if possible, to find out the advantages and disadvantages of the individual approaches,  as  well  as  the  sensitivity  of  detection required in various clinical settings. The purpose of this Review is not to judge the relative merits of the two approaches, or to imply that minimal residual disease assessment  is  a  proven  therapeutic  goal  in  multiple myeloma,  but  to  provide  clear  criteria  that  can  be uniformly  applied  to  and  validated  in  future  clinical trials and studies. 虽然目前还没有任何一种检测方法能完全满足所有这些理想标准，但下一代测序和下一代流式细胞技术可以满足其中的大部分标准，并可转化为一种先进的平台，在不同机构和国家统一应用。据报道，下一代测序和下一代流式细胞技术具有不同程度的灵敏度。这两种方法都能在 10⁵-10⁶ 个细胞中检测出一个多发性骨髓瘤细胞。我们强烈鼓励在可能的情况下将这两种方法纳入前瞻性试验，以找出各自方法的优缺点，以及不同临床环境下所需的检测灵敏度。本综述的目的不是评判这两种方法的相对优劣，也不是暗示最小残留病灶评估是多发性骨髓瘤行之有效的治疗目标，而是提供可在未来临床试验和研究中统一应用和验证的明确标准。

Defining a bone marrow minimal residual disease-negative response category 确定骨髓极小残留病阴性反应类别

The current proposal builds on the existing International Myeloma  Working  Group  response  criteria  by  adding additional   assessment  for  the   detection   of  minimal residual    disease    in   the   bone    marrow    (table    4). A comprehensive approach to detect very small amounts of disease both inside and outside of the marrow space will require  a  panel  of  tests  assessing  different  tumour compartments and probably use different technologies. However, these additional evaluation methods will require more data to show they complement existing methods and their  clinical  usefulness,  to  support  their  inclusion  in future iterations of International Myeloma Working Group response criteria. Furthermore, the added criteria should allow researchers to define a response state that reflects a higher degree of tumour eradication than is possible with the current definition of complete response or stringent complete response. At this time, we recommend the use of next-generation sequencing or next-generation flow for the detection of minimal residual disease in the bone marrow based on the availability of the two techniques at each centre and the feasibility for individual clinical trials. 目前的建议以现有的国际骨髓瘤工作组反应标准为基础，增加了检测骨髓内极小残留疾病的额外评估（表4）。要全面检测骨髓腔内外的极少量疾病，需要对不同的肿瘤分区进行一系列检测评估，并可能使用不同的技术。不过，这些额外的评估方法需要更多的数据来证明其对现有方法的补充作用及其临床实用性，以支持将其纳入国际骨髓瘤工作组反应标准的未来迭代中。此外，新增的标准应能让研究人员确定一种反应状态，这种状态所反映的肿瘤根除程度要高于目前完全反应或严格完全反应的定义。目前，我们建议使用下一代测序或下一代流式细胞技术来检测骨髓中的最小残留病，但这取决于各中心是否具备这两种技术以及个别临床试验的可行性。

 Accordingly, when minimal residual disease results are reported, the assessment should bequalified by the method(s) used (flow minimal residual disease-negative or sequencing minimal residual disease-negative), and the level of sensitivity (eg, one in 10⁵ or one in 10⁶ cells). Several  ongoing  studies  are  simultaneously  testing both methods, which will allow researchers to identify whether both techniques perform equally or whether one  approach  is  better  than  the  other.  Alternatively, both  methods  might  be  required  given  the  evolving clonal diversity of plasma cells.  Further work  should be  done  to  establish  whether  potentially  emerging alternative   cytometric   and   sequencing   techniques can  be  standardised  and  directly  compared  with  the next-generation  flow   EuroFlow  and  next-generation sequencing LymphoSIGHT methods.

因此，在报告极小残留病结果时，应根据所使用的方法（流式极小残留病阴性或测序极小残留病阴性）和灵敏度水平（例如，10⁵个细胞中一个或 10⁶个细胞中一个）来进行评估。目前正在进行的几项研究正在同时测试这两种方法，这将使研究人员能够确定这两种技术的性能是否相同，或者一种方法是否优于另一种方法。 另外，鉴于浆细胞克隆多样性的不断发展，两种方法可能都需要。 应进一步开展工作，确定潜在的新兴替代性细胞检测和测序技术是否可以标准化，并与下一代流式细胞检测EuroFlow和下一代测序LymphoSIGHT方法进行直接比较。

Detection of extramedullary disease 髓外疾病的检测

Present approaches for the detection and measurement of tumour burden after therapy rely on bone marrow assessment.  However,  bone  marrow  involvement  in multiple myeloma can be heterogeneous, thus increasing the likelihood  of a false-negative  assessment.  Further- more, such involvement does not allow detection of the disease outside the bone marrow. Extramedullary disease is increasingly seen in the clinic as a result of sensitive imaging studies and extended survival of patients with multiple myeloma. The estimated incidence of clinically detected  extramedullary  disease  among  a  cohort  of patients seen over a 10-year time period was 9%, with high-risk patients having a high risk of extramedullary disease later in the disease course. In the future, these rates might increase as increasingly sensitive imaging technologies  and novel biomarkers  are used to  detect minimal   residual   disease,    and   as   overall    survival continues to increase. This factor is of great relevance when response and disease progression are redefined, and particularly relevant when eradication of minimal residual  disease  is  redefined  in  the  context  of  new therapies. To ensure complete eradication of the tumour, assessment of the extramedullary compartment will be important as part of the disease assessment in multiple myeloma, particularly for defining high-quality complete response. 目前检测和测量治疗后肿瘤负荷的方法依赖于骨髓评估。 然而，多发性骨髓瘤的骨髓受累可能是异质性的，从而增加了假阴性评估的可能性。 此外，骨髓受累也无法检测到骨髓外的疾病。由于敏感的成像研究和多发性骨髓瘤患者生存期的延长，髓外疾病越来越多地出现在临床上。据估计，10 年间一组患者中临床发现的髓外疾病发生率为 9%，高危患者在病程后期发生髓外疾病的风险较高。未来，随着越来越灵敏的成像技术和新型生物标志物被用于检测极小残留病，以及总生存率的不断提高，这些比率可能会增加。在重新定义反应和疾病进展时，这一因素具有重要意义，而在新疗法背景下重新定义根除极小残留病时，这一因素尤为重要。为确保彻底根除肿瘤，髓外区的评估将成为多发性骨髓瘤疾病评估的重要组成部分，尤其是在确定高质量完全缓解时。

 PET/CT scans PET/CT 扫描

Improved imaging techniques have shown that multiple myeloma   can  be  heterogeneous   in   its   distribution pattern.   For  example,  the  pattern  of  bone  marrow infiltration by malignant plasma cells can vary between patients and within the same patient. In addition, studies suggest that up to 10% of patients (probably higher with more sensitive technologies) have extramedullary disease with the involvement of soft tissue or major organs at the time of diagnosis and suggest that a high proportion of patients  have  these  findings  at  the  time  of  disease relapse. 经改进的成像技术显示，多发性骨髓瘤的分布模式可以是异质性的。 例如，恶性浆细胞在骨髓中的浸润模式在不同患者之间和同一患者内部都会有所不同。此外，研究表明，多达10%的患者（采用更敏感技术的患者比例可能更高）在诊断时患有累及软组织或主要器官的髓外疾病，并表明有很高比例的患者在疾病复发时有这些表现。

¹⁸F-fluorodeoxyglucose  (¹⁸F-FDG)  PET  is  a  powerful tool to assess tumour metabolic activity and the effect of therapy  on  tumour-cell  metabolism.  Multiple  studies support  the  notion  that  the  detection  of  PET-positive lesions has  prognostic value  in  patients with  multiple myeloma   at   diagnosis   and   at   time   of   relapse. In addition to metabolic assessment, the low-dose CT that is typically done for localisation along with ¹⁸F-FDG PET is  a  sensitive  screen  for  multiple  myeloma-associated bone  disease.  In  an  initial  study, complete  ¹⁸F-FDG suppression in the focal lesions before first transplantation was  associated with better  survival  outcomes.  Another study  showed that persistent ¹⁸F-FDG avidity 7 days after the initiation of therapy was associated with worse survival outcomes  and  was  independent  of  other  prognostic factors. In an Italian study, PET/CT was performed at diagnosis,  after  thalidomide-dexamethasone  induction therapy  and  after  double  ASCT,  in  192  patients  newly diagnosed   with    multiple   myeloma.    Persistence   of maximum  standardised  uptake values  (SUVmax)  greater than 4·2 after induction therapy predicted an early relapse, and 4-year progression-free survival and overall survival was better  for those patients with  negative  PET/CT  at day 100 post-ASCT. PET/CT was negative in 125 (65%) of 192 patients 3 months post-ASCT. 4-year progression-free survival  was  47%  and  overall  survival  was  79%  for PET/CT-negative patients, compared with 32% (p=0·02) progression-free   survival   and   66%    (p=0·02)   overall survival, for PET/CT-positive patients. The Italian group presented updated results from their  study,  including 282 patients with newly diagnosed multiple myeloma who had  PET  imaging  at  baseline.  After  treatment,  PET negativity  was  achieved  in  132   (70%)  of  189  patients, whereas conventionally defined complete response was achieved in 104 (55%) patients. Among the proportion of patients who achieved a complete response, 30 (29%) had positive  PET  scans  and lower progression-free  survival (median 44 months vs 84 months, p=0·0009) and overall survival  (5-year  estimate  of  70%  vs  90%,  p=0·0032) compared with those with a positive PET/CT. In this study, persistence of SUVmax  higher than 4·2 was the only factor independently associated with skeletal progression in the absence of conventional measures of disease progression. The IFM2009 trial86  showed a clear value for PET imaging in   response   assessment   in   myeloma.   In  this  trial, 134 patients  had  a  PET/CT  scan  and  MRI  (spine  and pelvis) at study entry, at 3 months, and before maintenance therapy.  MRI  of the  spine  and  pelvis  and  whole-body PET/CT were equally effective in the detection of bone involvement in symptomatic patients at diagnosis. The median number of focal lesions detected by PET/CT was three (range 0 to more than ten lesions), with a median SUVmax    of 4·1  (range  1·5–28·4).  Normalisation  of the PET/CT was noted in 43 (32%) of 134 patients after three cycles   of  induction,   and   this   group   had   improved progression-free   survival   compared   with   those   with positive   PET/CT;   however,    overall   survival    did   not significantly  differ.  Normalisation  of  PET was  seen  in 83   (62%)   of   134   patients   before   maintenance   and progression-free   survival    and   overall   survival   were improved. The results of this study show the value of PET scanning in assessing treatment response during therapy inpatients with multiple myeloma. ¹⁸F-脱氧葡萄糖（¹⁸F-FDG）正电子发射计算机断层显像（PET）是评估肿瘤代谢活动和治疗对肿瘤细胞代谢影响的有力工具。 多项研究证实，PET阳性病灶的检测对多发性骨髓瘤患者的诊断和复发具有预后价值。除了新陈代谢评估外，通常与¹⁸F-FDG正电子发射计算机断层扫描一起进行的低剂量计算机断层扫描也是对多发性骨髓瘤相关骨病的敏感筛查。 在一项初步研究中，首次移植前病灶中的¹⁸F-FDG完全抑制与更好的生存结果相关。 另一项研究表明，治疗开始 7 天后，¹⁸F-FDG 阳性持续存在与较差的生存预后有关，且与其他预后因素无关。意大利的一项研究对192名新确诊的多发性骨髓瘤患者在诊断时、沙利度胺-地塞米松诱导治疗后和双份ASCT后进行了PET/CT检查。 诱导治疗后最大标准化摄取值（SUVmax）持续大于4-2预示着早期复发，ASCT后第100天PET/CT阴性的患者4年无进展生存率和总生存率更高。192名患者中，有125人（65%）在ASCT后3个月PET/CT呈阴性。PET/CT阴性患者的4年无进展生存率为47%，总生存率为79%，而PET/CT阳性患者的无进展生存率为32%（P=0-02），总生存率为66%（P=0-02）。意大利研究小组展示了他们研究的最新结果，其中包括282名基线PET成像的新诊断多发性骨髓瘤患者。 治疗后，189 名患者中有 132 人（70%）的 PET 呈阴性，104 人（55%）的 PET 呈完全反应。在获得完全应答的患者中，有 30 人（29%）PET 扫描呈阳性，与 PET/CT 呈阳性的患者相比，无进展生存期（中位 44 个月 vs 84 个月，p=0-0009）和总生存期（5 年估计为 70% vs 90%，p=0-0032）较低。在这项研究中，SUVmax持续高于4-2是在没有常规疾病进展指标的情况下唯一与骨骼进展独立相关的因素。IFM2009 试验86 显示了 PET 成像在骨髓瘤反应评估中的明确价值。 在该试验中，134 名患者在研究开始时、3 个月时和维持治疗前接受了 PET/CT 扫描和 MRI（脊柱和骨盆）检查。 脊柱和骨盆核磁共振成像与全身正电子发射计算机断层显像（PET/CT）在诊断时发现有症状患者的骨受累情况方面同样有效。PET/CT发现的病灶中位数为3个（范围从0到10多个病灶），中位SUVmax为4-1（范围1-5-28-4）。 134例患者中，有43例（32%）在三个周期的诱导治疗后PET/CT恢复正常，与PET/CT阳性的患者相比，这部分患者的无进展生存期有所改善；但是，总生存期没有显著差异。 134 例患者中有 83 例（62%）在维持治疗前 PET 恢复正常，无进展生存期和总生存期均有所改善。这项研究结果表明，PET扫描在评估多发性骨髓瘤患者治疗期间的治疗反应方面具有重要价值。

MRI

MRI examination is a sensitive method to detect bone marrow infiltration by multiple myeloma cells before bone    destruction    is    present    and    detectable    by conventional  radiographs.  The  role  of  MRI—both limited to the spine and whole-body approaches—has been studied extensively in the setting of symptomatic and  asymptomatic  patients  with  multiple  myeloma. Walker  and  colleagues  studied   611  patients  given different total therapy protocols, 452 (74%) of whom had focal  lesions  detected  by  baseline  MRI  that  correlate with known  prognostic  factors  in  multiple  myeloma. Hillengass   and   colleagues compared   conventional treatment   response   in   100   patients   with   multiple myeloma with whole-body MRI before and after ASCT. Good   concordance   was   noted   between   serological response  and  changes  in  imaging.  In this  study, the number of focal lesions present on post-therapy  MRI was informative for survival outcomes. Data from the IFM2009  trial86     demonstrated  equivalent  efficacy  for MRI  and   PET  in  the  detection  of  bone  lesions  at diagnosis.  MRI  normalisation  was  noted  in  a  small number  of  patients  (four  [3%]  after  three  cycles  of induction and  15  [11%] before maintenance), and did not translate into any improvement in progression-free survival or overall survival in this study. The usefulness of MRI  for  the  assessment  of residual  disease  after therapy remains unclear at this time due to the lack of sufficient data. 核磁共振成像检查是一种灵敏的方法，可在骨质破坏出现之前检测到多发性骨髓瘤细胞对骨髓的浸润，而传统的放射线检查则无法检测到这种浸润。 在有症状和无症状的多发性骨髓瘤患者中，核磁共振成像（仅限于脊柱和全身）的作用已得到广泛研究。沃克及其同事研究了611名接受不同总体治疗方案的患者，其中452人（74%）在基线磁共振成像中发现了病灶，这些病灶与多发性骨髓瘤的已知预后因素相关。Hillengass及其同事对100名多发性骨髓瘤患者在ASCT前后的常规治疗反应与全身磁共振成像进行了比较。结果表明，血清学反应与影像学变化之间具有良好的一致性。 在这项研究中，治疗后核磁共振成像显示的病灶数量对生存结果具有参考价值。IFM2009 试验86 的数据显示，MRI 和 PET 在诊断时检测骨病变的效果相当。 少数患者（三个诱导周期后 4 例 [3%]，维持治疗前 15 例 [11%]）的 MRI 恢复正常，但在本研究中并未转化为无进展生存期或总生存期的改善。由于缺乏足够的数据，核磁共振成像在评估治疗后残留疾病方面的作用目前仍不明确。

Defining an imaging response category 确定成像反应类别

Improvement  of the  limits  of disease  detection  with available  technologies  will  also  require  evaluation  of disease outside the bone marrow.  Present data favour the  use   of  ¹⁸F-FDG  PET.  One  study  examined  the diagnostic efficacy of whole-body MRI versus ¹⁸F-FDG PET  in  31  patients  after  stem-cell  transplantation.  In this study, 104 lesions were detected in 21 patients: PET/CT  had  a  lower  sensitivity  than  MRI  (50·0%  vs 80·0%), a higher specificity (85·7% vs 38·1%), a higher positive  predictive  value   (62·5%  vs   38·1%),  a  lower negative  predictive  value  (78·3%  vs  80·0%),  and  was more accurate overall for the determination of remission status  (74·2% vs  51·6%). While  some  studies  suggest that MRI is more sensitive in picking up lesions at the time  of initial  evaluation,  ¹⁸F-FDG  PET  has  distinct advantages for follow-up evaluation. Metabolic changes on ¹⁸F-FDG PET can detect early responses, but MRI responses   are   usually    delayed   as    marrow   signal abnormalities can take a long time to resolve depending on  the  size  of  the  lesion.84,91,92      MRI  also  has  a  low specificity in the differentiation of viable disease from bone   remodelling   compared   with   ¹⁸F-FDG   PET.93,94  However,   for   minimal   residual   disease   monitoring (in which ¹⁸F-FDG uptake is important rather than lytic bone    lesion    detection),    both    false-negative    and false-positive   results    (in   case    of   other   coexisting infectious or inflammatory processes) maybe seen. Data from the IFM2009 trial have shed some light on the additive  value   of  imaging-based   and   marrow-based assessments of minimal residual disease. Among the 134 patients assessed by PET at various stages of therapy, 利用现有技术提高疾病检测的极限还需要对骨髓外的疾病进行评估。 目前的数据支持使用¹⁸F-FDG PET。 一项研究对干细胞移植后的31名患者进行了全身核磁共振成像与¹⁸F-FDG PET的诊断效果对比。 在这项研究中，21 名患者共发现了 104 个病灶：PET/CT的灵敏度低于核磁共振成像（50-0%对80-0%），特异性较高（85-7%对38-1%），阳性预测值较高（62-5%对38-1%），阴性预测值较低（78-3%对80-0%），总体而言，对缓解状态的判断更为准确（74-2%对51-6%）。一些研究表明，核磁共振成像在初次评估时发现病灶的灵敏度更高，但¹⁸F-FDG PET 在后续评估中具有明显优势。¹⁸F-FDG正电子发射计算机断层显像的代谢变化可检测出早期反应，但核磁共振成像的反应通常会延迟，因为骨髓信号异常需要很长时间才能消除，这取决于病变的大小84,91,92。与¹⁸F-FDG正电子发射计算机断层显像相比，核磁共振成像在区分存活疾病和骨重塑方面的特异性较低93,94。 93,94 但是，对于极小残留病监测（¹⁸F-FDG 摄取比溶解性骨病变检测更重要），可能会出现假阴性和假阳性结果（如果同时存在其他感染或炎症过程）。IFM2009 试验的数据揭示了基于成像和基于骨髓的最小残留病评估的附加价值。在不同治疗阶段接受 PET 评估的 134 名患者中：

results of minimal residual disease by flow cytometry were available in 86 patients. Progression-free survival was improved for the 41 patients who had negative bone marrow PET results compared with those patients who had positive results using either or both methods. 86例患者通过血流细胞学检测获得了最小残留病灶的结果。与使用两种方法中的任何一种得出阳性结果的患者相比，骨髓 PET 阴性结果的 41 名患者的无进展生存期有所改善。

The data available from these studies show an inferior outcome for patients with positive  PET  scans even in those  who  achieved  deep  responses,  highlighting  the relevance  of this  assessment method in patients with myeloma.  PET/CT  has become  standard  for  response assessment   in   lymphomas,   where   baseline   scans, interval scans during treatment, and end of treatment scans are integrated into the response criteria. A specific five-point   scoring    system   has    been   developed    to standardise the scoring of images to define response on serial scans (Deauville criteria). In the present criteria, we have defined the imaging response stringently as the disappearance of every area of increased tracer uptake found at baseline, or a preceding PET/CT; or a decrease to less than the mediastinal blood pool SUV; or a decrease to   less   than   that   of   surrounding   normal   tissue. These criteria are analogous to what has been used in lymphoma in which a complete metabolic response has been defined as a score of one or two on the five-point scale.  Response  assessments  should  be  conservative, because myeloma remains incurable and use of these criteria in prospective clinical trials should not lead to the undertreatment of patients. Future prospective trials will allow fine tuning of the cutoffs used for defining absence of disease on PET imaging. 这些研究的数据显示，PET 扫描呈阳性的患者即使获得深度反应，疗效也较差，这突出表明了这种评估方法对骨髓瘤患者的意义。 PET/CT 已成为淋巴瘤反应评估的标准，基线扫描、治疗期间的间隔扫描和治疗结束扫描都纳入了反应标准。目前已开发出一套具体的五点评分系统，对图像进行标准化评分，以确定连续扫描的反应（多维尔标准）。在本标准中，我们将成像反应严格定义为：基线或之前的 PET/CT 发现的示踪剂摄取增加区域全部消失；或下降至低于纵隔血池 SUV；或下降至低于周围正常组织的 SUV。这些标准与淋巴瘤的标准类似，淋巴瘤的完全代谢反应被定义为五级评分中的一级或二级。 由于骨髓瘤仍是不治之症，在前瞻性临床试验中使用这些标准不应导致对患者的治疗不足，因此对反应的评估应持保守态度。未来的前瞻性试验将允许对用于确定 PET 成像无疾病的切线进行调整。

Many questions remain incompletely answered—eg, how  many  flow  minimal  residual  disease-negative  or molecular  minimal  residual  disease-negative  patients are   imaging   positive?    In   which   patients    should clinicians  be  particularly  aware  of  the  potential  for extramedullary disease? Do investigators need the same imaging technique at baseline and after treatment to evaluate    metabolic    response?     Should    treatment (consolidation/maintenance)  be  tailored  on  imaging- defined   minimal   residual   disease?    For   example, extramedullary relapses are likely even among minimal residual disease-negative patients after ASCT suggesting that,  at  least  for  this  particular  therapeutic  strategy, response  assessment  might  benefit  from  combined medullary  and  extramedullary  (PET/CT)  measure  of minimal   residual   disease.    In   turn,   standardised interpretation    of   imaging   techniques   remains    a challenge.   Several   attempts   to   standardise   criteria for   PET/CT   imaging   definitions   and   the   use   of semi-quantitative SUV evaluations are now ongoing to consolidate the use of this technique  as a prognostic tool. New imaging technologies such as PET/MRI have been introduced.  PET in  combination with  MRI is  a novel and promising method, in which the PET detects active focal lesions, while the MRI shows the location of the lesions and provides information on myeloma-cell infiltration of the bone marrow. By substitution of the 许多问题仍未得到完整解答--例如，有多少血流极小残留病阴性或分子极小残留病阴性患者的影像学结果呈阳性？ 临床医生应特别注意哪些患者可能存在髓外疾病？研究者是否需要在基线和治疗后使用相同的成像技术来评估代谢反应？ 治疗（巩固/维持）是否应根据影像学确定的最小残留病来定制？ 例如，即使在 ASCT 治疗后极小残留病阴性的患者中也可能出现髓外复发，这表明，至少对于这种特殊的治疗策略而言，联合髓内和髓外 （PET/CT）测量极小残留病可能有利于反应评估。 反过来，成像技术的标准化解释仍然是一项挑战。 目前正在尝试对 PET/CT 成像的定义和半定量 SUV 评估的使用标准进行标准化，以巩固该技术作为预后工具的应用。PET/MRI 等新的成像技术已经问世。 正电子发射计算机断层显像（PET）与核磁共振成像（MRI）相结合是一种新颖而有前途的方法，正电子发射计算机断层显像（PET）可检测活跃的病灶，而核磁共振成像（MRI）可显示病灶的位置，并提供骨髓瘤细胞浸润骨髓的信息。通过替代

CT component in PET/CT, MRI not only provides the anatomical  localisation,  but  also  brings  two  active modalities  into  a  single  study,  with  relatively  short acquisition time without compromising on the imaging quality and avoiding the radiation exposure associated with   CT.   The   results   of  a   study   that   compared PET/CT    and    functional    MRI—namely,    diffusion- weighted imaging—as a whole-body protocol in a small group of patients with multiple myeloma, showed that diffusion-weighted  imaging  is  superior  in  detecting focal  and  diffuse  infiltration  of  the  bone  marrow. Further   studies   should   investigate   which   imaging technique or which combination brings the most final benefit  for  patients with  multiple  myeloma  in  initial investigations and response assessment. 与 PET/CT 中的 CT 部分相比，MRI 不仅能提供解剖定位，还能将两种有源模式整合到一项研究中，而且采集时间相对较短，不会影响成像质量，还能避免与 CT 相关的辐射暴露。 一项研究比较了 PET/CT 和功能磁共振成像（即二重灌注加权成像），并将其作为一小组多发性骨髓瘤患者的全身方案，结果显示二重灌注加权成像在检测骨髓的局灶性和二重灌注方面更具优势。进一步的研究应探讨哪种成像技术或哪种成像技术的组合能为多发性骨髓瘤患者的初步检查和反应评估带来最大益处。

Special considerations based on therapy 基于治疗的特殊考虑

Monoclonal  antibodies  are  a  promising  area  for  the treatment  of  multiple  myeloma,  and  several  will  be available in the clinic in the future. Use of monoclonal antibodies  can  present  unique  challenges  for  clinical response   assessment   techniques.   These   challenges include   interference   with   the    monoclonal   protein assessment   on    serum   protein   electrophoresis,   or immunofixation,  and  with  MFC-based  assessment  of monoclonal plasma cells in the bone marrow aspirates. 单克隆抗体是治疗多发性骨髓瘤的一个前景广阔的领域，未来将有多种单克隆抗体应用于临床。使用单克隆抗体会给临床反应评估技术带来独特的挑战。 这些挑战包括干扰血清蛋白电泳中的单克隆蛋白评估或免疫吸附，以及干扰基于 MFC 的骨髓抽吸物中单克隆浆细胞评估。

The monoclonal antibodies that have been approved, as well as those in clinical development, can be detected on  the  immunofixation  assays  currently  used  in  the clinic for the detection of small amounts of monoclonal protein.  This  factor  is  important  because  complete response  is  defined  as  the  complete  disappearance of  the   monoclonal   protein   on   serum   and   urine immunofixation.    When    the    infused    monoclonal antibody  shares  the  same  isotype  as  the  monoclonal multiple myeloma protein, low levels of the therapeutic antibody  can  lead  to  a  false-positive  immunofixation result, potentially under-reporting the drug’s depth of response. Anti-idiotype  antibodies  that  bind  the offending drug and alter its migration out of the range of the  endogenous  M-protein,  allow  confirmation  of interference  on  serum  immunofixation  and  protein electrophoresis, and assays based on this strategy are being   developed   for   mitigation   of  this   problem.102  To   help  with   this   issue,   mass   spectrometry-based techniques that enable the discrimination of different proteins based on their masses are being developed. Although confirmation of serological complete response might   not   alter   treatment   decisions   in   day-to-day practice,  these  endpoints  are  key  in  the  clinical  trial setting; therefore, reflex testing to distinguish between the monoclonal protein and the therapeutic antibody in patients who are immunofixation positive only should be mandatory in clinical trial settings. 已获批准的单克隆抗体以及正在临床开发的单克隆抗体，都可以通过目前临床上用于检测少量单克隆蛋白的免疫吸附试验检测出来。 这一因素非常重要，因为完全应答是指血清和尿液免疫吸附中的单克隆蛋白完全消失。 当注入的单克隆抗体与单克隆多发性骨髓瘤蛋白具有相同的同种型时，低水平的治疗性抗体会导致免疫测定结果呈假阳性，从而可能低估药物的深度反应。抗异型抗体可结合末端药物并改变其迁移，使其脱离内源性 M 蛋白的范围，从而确定对血清免疫吸附和蛋白质电泳的干扰，目前正在开发基于这一策略的检测方法，以缓解这一问题102。尽管血清学完全反应的确认可能不会改变日常治疗决策，但这些终点在临床试验中却非常关键；因此，在临床试验中必须对免疫反应阳性的患者进行再检测，以区分单克隆蛋白和治疗抗体。

The therapeutic approach taken can also have an effect on minimal residual disease testing by MFC. CD38 is a critical  surface marker that is extensively used for the identification of plasma cells by flow cytometry, and the use of anti-CD38 antibodies can potentially interfere with the flow cytometry-based assay. To this end, specific CD38 antibody  clones  or  reagents,  together  with  the  most sensitive CD138-fluorochrome conjugates, such as those validated  and  incorporated  in  the  current  EuroFlow  2 tube  eight-colour panel  (eg,  the  CD38  multiclone  and CD138-BV421  reagents),  will   allow  for   a  treatment- independent  minimal  residual  disease  assay  with  the greatest  sensitivity  and  specificity.  By  contrast,  next- generation  sequencing  is  not  affected  by  monoclonal antibody-based  treatments.  Other  promising  therapies that are currently going through clinical trials include chimeric antigen receptor T cells, which can influence the   immune-cell   types   and   may   require   additional strategies that are yet to bedefined. 所采取的治疗方法也会对MFC检测的最小残留病产生影响。CD38 是一种重要的表面标志物，被广泛用于通过流式细胞术鉴定浆细胞，而使用抗 CD38 抗体可能会干扰基于流式细胞术的检测。为此，特异性 CD38 抗体克隆或试剂与最灵敏的 CD138-荧光共轭物（如目前 EuroFlow 2 管式八色板（如 CD38 多克隆试剂和 CD138-BV421 试剂）中已验证和纳入的那些试剂）一起使用，将能以最高的灵敏度和特异性进行独立于治疗的最小残留病检测。 相比之下，新一代测序不受单克隆抗体疗法的影响。 目前正在进行临床试验的其他有前途的疗法包括嵌合抗原受体T细胞，这种疗法会影响免疫细胞类型，可能需要采取其他策略，但这些策略尚有待确定。

Updated consensus response criteria 更新共识响应标准

The  present  iteration  of  the   International   Myeloma Working  Group  consensus  response  criteria  has  been crucial in light of the progress witnessed over the past decade in the development of new drugs and treatment approaches, including high-dose therapy, consolidation, and maintenance approaches. Ambiguities and nuances have become apparent in these criteria as they are used in multicentre  clinical  trials  performed  across  different geographical  regions,  with  highly  effective  treatment regimens, including drugs with new methods of action. Uniform response criteria should not only be used across all  clinical  trials,  but  they   should  also  be  uniformly interpreted and applied. To provide a clear approach to the application of the response criteria, we have incorporated many  practical  clarifications  in  the  current  consensus criteria (table 4). We hope that this will serve as a practical guide  for  investigators  and  pharmaceutical  companies involved in clinical trials for multiple myeloma. 过去十年间，新药和治疗方法（包括大剂量治疗、巩固治疗和维持治疗）的开发取得了长足进步，因此，国际骨髓瘤工作组共识反应标准的当前版本至关重要。由于这些标准被用于不同地理区域的多中心临床试验，并采用高效的治疗方案，包括具有新作用方法的药物，因此其模糊性和细微差别已变得显而易见。所有临床试验不仅应使用统一的反应标准，还应统一解释和应用这些标准。为了提供明确的反应标准应用方法，我们在当前的共识标准中纳入了许多实用的分类（表 4）。我们希望这将成为参与多发性骨髓瘤临床试验的研究人员和制药公司的实用指南。

Baseline measurements and required testing during follow-up

In addition to tumour burden-based response assess- ment,   other   laboratory   measurements   have   been incorporated into the current response criteria to define a category of clinical progression. This categorisation is  particularly  important  as  oncologists  increasingly encounter   oligo-secretory   disease   or   non-secretory disease  in  patients  who  had  measurable  levels  of monoclonal protein at the time of diagnosis. While we believe that this situation reflects clonal evolution of the multiple   myeloma   cells,   the   precise   mechanisms remain poorly understood. Thus, guidelines that reflect functional consequences of disease progression such as haemoglobin, renal function, and serum calcium need to  be  followed  closely.  Table  5  defines  the  required baseline and ongoing testing in patients with multiple myeloma that are key for appropriate application of the consensus  criteria.  The  panel  provides  guidance  on commonly observed situations in patients enrolled in clinical trials.  Definitions for time-to-event endpoints can be found in a previous publication.14  We propose to redefine    disease-free    progression    using    minimal residual   disease    rather   than    complete   response: duration  from  the  start  of minimal  residual  disease negativity  to  the  time  of  reappearance  of  minimal residual disease. In this definition, disease-free survival only  applies  to  patients  who  are  minimal  residual disease-negative. 除了基于肿瘤负荷的反应评估外，其他实验室测量结果也被纳入了目前的反应标准，以确定临床进展的类别。这种分类尤为重要，因为肿瘤学家越来越多地遇到在诊断时单克隆蛋白水平可测量的患者出现少分泌性疾病或非分泌性疾病的情况。虽然我们认为这种情况反映了多发性骨髓瘤细胞的克隆进化，但对其确切机制仍知之甚少。因此，需要密切关注反映疾病进展的功能性后果（如血红蛋白、肾功能和血清钙）的指南。 表 5 列出了多发性骨髓瘤患者所需的基线检测和持续检测项目，这些项目是适当应用共识标准的关键。 专家小组就临床试验入组患者的常见情况提供了指导。 14 我们建议使用最小残留病而非完全反应来重新定义无病进展：从最小残留病阴性开始到最小残留病再次出现的持续时间。根据这一定义，无病生存期仅适用于极小残留病阴性的患者。

Future directions 未来方向

The  development  of  an  accurate  framework  for  the assessment  of minimal  residual  disease  is  a work  in progress  and this report is the first  and probably the most important step in that direction. Ongoing work will continue to define what level of minimal residual disease is clinically relevant and when it should be evaluated. Specific aspects of disease biology will also need to be incorporated  into  future   definitions  of  the  minimal residual  disease  state  (eg,  identification  of  minimal residual disease-positive patients who will nevertheless experience long-term survival). 为评估极小残留病制定一个准确的框架是一项正在进行的工作，本报告是朝着这个方向迈出的第一步，也可能是最重要的一步。我们将继续开展工作，确定何种程度的极小残留病具有临床相关性，以及何时应对其进行评估。疾病生物学的具体方面也需要纳入未来对极小残留病状态的定义中（例如，确定极小残留病阳性但仍能长期存活的患者）。

Detection of minimal residual disease in blood

Clonal plasma cells in multiple myeloma  are typically restricted to the bone marrow, although small numbers can be detected by sensitive approaches in the peripheral blood of most patients with newly diagnosed or relapsed multiple myeloma. In both newly diagnosed and relapsed disease,the presence of circulating tumour cells has been associated  with   shorter  progression-free   survival  and inferior overall survival. In a study of 647 consecutive patients with previously treated multiple myeloma who had   their   peripheral   blood   evaluated   for   multiple myeloma plasma cells by MFC, none of the patients who achieved  a  complete  response  had  circulating  plasma cells at the time of initial evaluation at the  study  site compared with 62 (9·6%) of 647 patients with relapsed disease. Demonstration of absence of multiple myeloma cells in circulation may be important for all patients with multiple  myeloma,  particularly  for  those  with  large numbers   of   circulating   cells   at   initial   evaluation. DNA-sequencing  methods  have  also  been  applied  to detect small numbers of circulating tumour cells in the peripheral blood. In one study, minimal residual disease was  assessed  in  42  patients  undergoing  ASCT  using circulating DNA in the peripheral blood that was analysed by ASO-qPCR to identify rearranged IGH genes.  Even though the minimal residual disease level in peripheral blood  samples  was  significantly  lower  than  in  bone marrow   samples,  patients  with  negative  ASO-qPCR results  3  months  after  ASCT  had  a  longer  event-free survival (median 15 months vs 4 months; p=0·004) and longer overall survival (median 52 months vs 17 months; p=0·03). Importantly, sequential monitoring of clonotypic cells in peripheral blood allowed the early identification of disease relapse. Another study used a sequencing-based method to identify multiple myeloma cells in peripheral blood samples, and was able to detect clones at less than one  in  a  million  leucocytes  (0·0001%).  The  authors detected multiple myeloma cells in the peripheral blood in  44  (96%)  of  46  patients.  Although  there  was  a correlation  between  multiple  myeloma  clone  levels  in paired  bone  marrow  and  peripheral  blood  samples, almost all patients investigated in these studies did not achieve a complete response. Prospective studies should examine the true prognostic value  of the  detection  of multiple myeloma cells in the circulation of patients who achieve a complete response and compare these results to those obtained in paired bone marrow samples before these methods can be adopted. 多发性骨髓瘤中的克隆性浆细胞通常局限于骨髓中，但大多数新诊断或复发的多发性骨髓瘤患者的外周血中都能通过灵敏的方法检测到少量浆细胞。在新诊断和复发的疾病中，循环肿瘤细胞的存在与较短的无进展生存期和较差的总生存期有关。一项针对 647 名曾接受过治疗的连续多发性骨髓瘤患者的研究通过 MFC 对其外周血中的多发性骨髓瘤浆细胞进行了评估，结果显示，在研究地点进行初步评估时，获得完全应答的患者中没有一人有循环浆细胞，而在 647 名复发患者中，有 62 人（9-6%）有循环浆细胞。对于所有多发性骨髓瘤患者，尤其是在初次评估时有大量循环细胞的患者来说，证明循环中没有多发性骨髓瘤细胞可能非常重要。DNA 测序方法也被用于检测外周血中少量的循环肿瘤细胞。一项研究利用外周血中的循环 DNA 评估了 42 名接受 ASCT 的患者的极小残留病，并通过 ASO-qPCR 分析确定了重排的 IGH 基因。 尽管外周血样本中的极小残留病水平明显低于骨髓样本，但 ASCT 3 个月后 ASO-qPCR 结果为阴性的患者无事件生存期（中位 15 个月 vs 4 个月；p=0-004）和总生存期（中位 52 个月 vs 17 个月；p=0-03）都更长。重要的是，通过对外周血中克隆型细胞的序列监测，可以及早发现疾病复发。另一项研究使用了一种基于测序的方法来识别外周血样本中的多发性骨髓瘤细胞，能够检测到小于百万分之一的白细胞克隆（0-0001%）。 作者在 46 名患者中的 44 人（96%）的外周血中检测到了多发性骨髓瘤细胞。 虽然配对骨髓和外周血样本中的多发性骨髓瘤克隆水平之间存在相关性，但这些研究中调查的几乎所有患者都没有获得完全应答。在采用这些方法之前，前瞻性研究应检查在获得完全应答的患者血液循环中检测多发性骨髓瘤细胞的真正预后价值，并将这些结果与配对骨髓样本中的结果进行比较。

Ongoing studies are examining the assessment of circulating tumour DNA as a sensitive measure of small amounts of residual cells. In addition to quantification, assessment of circulating tumour DNA levels could allow investigators    to    track    individual    tumour    clones. The sensitivity of blood for the evaluation of minimal residual disease remains unknown and the development of peripheral  blood-based  monitoring  should  be  the ultimate  goal  as  it  would  allow  for   serial  sampling without the trauma of repeated bone marrow aspirations. 目前正在进行的研究将循环肿瘤 DNA 评估作为少量残留细胞的灵敏测量方法。除了定量分析外，循环肿瘤 DNA 水平的评估还能让研究人员追踪单个肿瘤克隆。血液对评估微小残留病的敏感性仍是未知数，开发基于外周血的监测应是最终目标，因为它可以进行连续采样，而无需反复抽取骨髓。

Hevylite assay 赫维测定

In conjunction with the International Myeloma Working Group  response  criteria, the  ability to  quantitate  free immunoglobulin    light     chains     greatly    enhanced oncologists’ ability  to  detect  deeper  responses  and  to define stringent complete response. The development of antibodies  against  conjunction  epitopes  between  the light  and  heavy   chains  enables  the   quantitation  of specific    pairs    of   heavy/light    chains    (IgGκ/IgGλ, IgAκ/IgAλ,  and  IgMκ/IgMλ) in the  serum  and is the basis of the Hevylite assay (Binding Site, Birmingham, UK). The Hevylite assay provides information on both the  involved  immunoglobulin   (eg,  IgGκ  in  an  IgGκ patient) and the polyclonal non-involved pair (eg, IgGλ in  an  IgGκ  patient).  The  Hevylite  assay  is  useful  in patients with oligo-secretory disease and can overcome limitations   associated   with   monitoring   β-migrating monoclonal  IgA  by  electrophoresis. Studies  have  also indicated  a  role  of  the  Hevylite  assay  in  minimal residual disease assessment. Increased IgAκ/IgAλ and IgMκ/IgMλ  ratios   of  the   uninvolved   isotype   were associated    with    longer     progression-free    survival compared with normal ratios.114   This  probably reflects the degree of immune recovery post-ASCT, which could enhance  the  capacity  to  immunologically  control  the disease for longer. Unlike the other tests described so far, heavy/light chain ratios could reflect a functional consequence of minimal residual disease negativity on the  recovery  of  normal  B  cells  and  plasma  cells,  in addition to the quantitative estimate of residual disease. In most cases, responses assigned by the Hevylite assay have  shown  to  be   equivalent  to  those   assigned  by conventional    methods.    In    some    cases,    however, heavy/light chain ratios provided additional sensitivity. Ludwig  and  colleagues115     studied   sequential  sera  of 156   patients   with   IgG   or   IgA   multiple   myeloma comparing    the    heavy/light    chain    measurements with   conventional   assays   such   as   serum   protein electrophoresis,   immunofixation,   nephelometry,   and sFLC  tests.  When  both  heavy/light  chain  and  sFLC testing  were  applied  for  response  assessment,  clonal excess was noted in 14 (45%) of 31 patients who achieved a   complete   response.   The   heavy/light   chain   ratio indicated  the  presence  of  disease  in  eight   (26%)  of 31 patients who achieved a complete response and, in sequential    studies,   indicated    evolving   relapse   in three patients before immunofixation became positive. It  is  probable  that  the  test  not  only  allows  for  the detection of persistent secretory clones of plasma cells, but it is also an indicator of the normalisation of the immune  system,  suggesting  a  deeper  eradication  of the tumour clone and a lack of negative effect on the immune status. However, more data must be collected, particularly  among  patients  who  achieve  a  complete response, to allow conclusions to be drawn for the use of the Hevylite assay. 结合国际骨髓瘤工作组的反应标准，游离免疫球蛋白轻链的定量能力大大提高了肿瘤学家检测深度反应和确定严格的完全反应的能力。针对轻链和重链之间连接表位的抗体的开发使血清中重链/轻链（IgGκ/IgGλ、IgAκ/IgAλ和IgMκ/IgMλ）的特异性成对定量成为可能，这也是Hevylite检测法（英国伯明翰的Binding Site公司）的基础。Hevylite 检测法可提供相关免疫球蛋白（如 IgGκ 患者的 IgGκ）和多克隆非相关免疫球蛋白对（如 IgGκ 患者的 IgGλ）的信息。 Hevylite检测法适用于低分泌性疾病患者，可克服电泳法监测β迁移单克隆IgA的局限性。研究还表明，Hevylite 检测法在最小残留病评估中也能发挥作用。与正常比率相比，未受累同种型 IgAκ/IgAλ 和 IgMκ/IgMλ 比率的增加与更长的无进展生存期相关114。与迄今为止描述的其他检测方法不同，重链/轻链比值除了对残留疾病的定量估计外，还能反映最小残留疾病阴性对正常 B 细胞和浆细胞恢复的功能性影响。在大多数情况下，Hevylite 检测法得出的反应与传统方法得出的反应相当。 不过，在某些情况下，重链/轻链比值提供了额外的灵敏度。路德维希及其同事115 对 156 名 IgG 或 IgA 多发性骨髓瘤患者的连续血清进行了研究，并将重链/轻链测量结果与血清蛋白电泳、免疫吸附、浊度测定和 sFLC 检测等传统检测方法进行了比较。 在同时使用重链/轻链和 sFLC 检测进行反应评估时，31 例获得完全反应的患者中有 14 例（45%）出现了克隆过多。 在获得完全应答的 31 位患者中，有 8 位（26%）患者的重链/轻链比值显示存在疾病，在连续研究中，有 3 位患者的重链/轻链比值在免疫检测呈阳性之前就显示了疾病的复发。 很有可能的是，该试验不仅可以检测到持续存在的浆细胞分泌克隆，而且还是免疫系统正常化的一个指标，表明肿瘤克隆已被深入根除，对免疫状态没有负面影响。不过，必须收集更多的数据，尤其是在获得完全应答的患者中收集数据，才能为海威利特检测法的使用得出结论。

Timing and frequency of disease assessment 疾病评估的时间和频率

Disease biology plays a key role in the determination of the  degree  and  the  duration  of disease  control  after therapy.  For  example,  a  rapid  and  deep  response  is commonly  seen  in  patients  with  multiple  myeloma with  features  of high-risk  disease,  which  is  often— though not always—poorly sustained and followed by a rapid relapse. In a study by van Rhee and colleagues,116  sFLC levels were measured at baseline, within 7 days of starting  the  first  cycle,  and  before  both  the  second induction cycle and the first ASCT. Patients within the top  tercile  for  sFLC  reductions  from  baseline  until cycle  2  or  before  transplantation  (reflecting  either  a more  rapid  response  or  a  higher  tumour  burden  at presentation)  had  an  inferior  event-free  survival  and overall survival compared with the other two terciles. Barlogie   and   colleagues56       examined   the   effect   of complete    response    on    survival    among    patients undergoing   total   therapy   protocols.   The   authors observed that  patients who had  achieved  a  complete response   and   then   relapsed   had   inferior   survival compared with those who never achieved a complete response. These patients were more likely to have gene expression profile-defined high-risk multiple myeloma and more likely to present with other poor prognostic factors. It has become clear that these patients not only have  high-risk   features   at  baseline  but   also  have persistent minimal residual disease in the context of achieving  complete  response.  We  propose  that  the assessment of minimal residual disease kinetics over the disease course, rather than at a  single timepoint when  complete  response  is  first  documented,  could provide a more robust evaluation of disease control in patients  with  multiple  myeloma   after   achieving   a complete  response  or  stringent  complete  response. Conversely,   one   group117       proposed   that   a    small proportion of patients have a monoclonalgammopathy of unknown significance (MGUS)-like gene expression profile  signature,  and  they  experience  significantly better  outcomes  compared  with  the  vast  majority  of (non-MGUS-like)   patients   with   multiple   myeloma without necessarily increased complete response rates. More recently, a  Spanish group proposed that there are patients with multiple myeloma with an MGUS-like flow  cytometry  signature,  and  that  they  have  better outcomes  (estimated  60%  time  to  progression  and overall  survival  at   10  years)  independently  of  their complete response status. Altogether, in addition to the amount of tumour burden that persists after therapy, the genetic and epigenetic make-up of chemotherapy- resistant minimal residual disease cells might dictate the duration of survival. 疾病生物学在决定治疗后疾病控制的程度和持续时间方面起着关键作用。 例如，具有高危疾病特征的多发性骨髓瘤患者通常会出现快速而深刻的反应，但这种反应往往（尽管并非总是）持续性较差，随后会迅速复发。在 van Rhee 及其同事的一项研究中116 ，分别在基线、第一个周期开始后 7 天内、第二个诱导周期和第一次 ASCT 之前测量了 sFLC 水平。从基线到第二周期或移植前，sFLC降幅最高的三组患者（反映了更快的反应或发病时更高的肿瘤负荷）的无事件生存期和总生存期均低于其他两组。Barlogie及其同事56研究了完全反应对接受整体治疗方案的患者生存期的影响。 作者观察到，与从未获得完全应答的患者相比，获得完全应答后复发的患者生存率较低。这些患者更有可能患有基因表达明确的高危多发性骨髓瘤，也更有可能存在其他不良预后因素。现在已经很清楚，这些患者不仅在基线时具有高危特征，而且在获得完全应答的情况下还会有持续的极小残留病变。 我们建议，评估病程中的极小残留病变动力学，而不是在首次记录完全应答时的单一时间点进行评估，可以更可靠地评估多发性骨髓瘤患者在获得完全应答或严格完全应答后的疾病控制情况。与此相反，一个研究小组117 提出，一小部分患者具有意义不明的单克隆炎症（MGUS）样基因表达特征，与绝大多数（非 MGUS 样）多发性骨髓瘤患者相比，他们的预后明显更好，但完全应答率却不一定提高。最近，西班牙的一个研究小组提出，有一些多发性骨髓瘤患者具有MGUS样血流细胞学特征，他们的预后更好（估计60%的进展时间和10年的总生存率），与完全应答状态无关。 总之，除了治疗后持续存在的肿瘤负荷量外，对化疗产生抗药性的极小残留病细胞的基因和表观遗传构成也可能决定患者的生存期。

Conclusion 结论

The proposed guidelines form a framework for future investigation into minimal residual disease in multiple myeloma. Prospective studies are being incorporated in newly  designed  clinical  trials,  and  we  encourage  new studies  to  incorporate  (whenever  reasonable)  minimal residual disease monitoring by next-generation flow or next-generation  sequencing,  or  both. In   addition, existing archived samples from various clinical trials and different institutions are being evaluated for the validation of the  clinical usefulness  of minimal  residual  disease monitoring  as  a  predictive  variable.   In  view  of  the increasing incidence of extramedullary disease inpatients with multiple myeloma, the presence of extramedullary disease should be ruled out as part of minimal residual disease assessment. Ongoing studies are evaluating the role of a PET scan at the time of minimal residual disease assessment along with the previously mentioned testing, especially when minimal residual  disease negativity is achieved in the bone marrow. 拟议的指南为今后研究多发性骨髓瘤极小残留病提供了框架。前瞻性研究正被纳入新设计的临床试验中，我们鼓励新研究（在合理的情况下）通过下一代流式或下一代测序或两者同时进行最小残留病监测。此外，我们正在评估来自各种临床试验和不同机构的现有存档样本，以验证最小残留病监测作为预测变量的临床实用性。 鉴于多发性骨髓瘤患者髓外疾病的发病率越来越高，在评估极小残留病时应排除髓外疾病的存在。目前正在进行的研究正在评估在评估极小残留病时进行正电子发射计算机断层扫描以及前面提到的检测的作用，尤其是在骨髓中极小残留病阴性的情况下。

Finally, the use of heavy/light chain ratios might have an important role in the definition of a minimal residual disease-negative  state.  The  combination  of a  negative cell-based  assay,  negative   PET  scan,  and  a  normal heavy/light chain ratio probably represents a composite endpoint reflecting the eradication of tumour cells from all compartments and recovery of the normal plasma-cell population to the currently available level of detection. This aspect needs further study in prospective clinical trials and large retrospective datasets. Development of a blood-based assay, either testing for rare circulating cells or circulating tumour DNA, would be ideal, and ongoing work should be focused on developing these approaches.  As  more  sensitive  flow-based  assays  become  more commonplace, we anticipate that the stringent complete response criteria will be used less frequently and may eventually be dropped. This factor is particularly relevant, as the contribution of sFLC normalisation as part of the stringent complete response criteria has been challenged by data from the Intergroupe Francophone du Myélome group.  If indeed the usefulness of stringent complete response over complete response comes mostly from the lack of detectable plasma cells by less sensitive methods, use of minimal residual disease methods will make this criterion obsolete. Another area of active investigation has been the  substitution  of sFLC  measurements  for 24-h urine measurements. While this substitution would greatly reduce the burden for patients and physicians, no definitive  data  support  this  change  at  this  time.118–120 The most important question that this approach raises is the  effect  of the  minimal  residual  disease  results  on decisions regarding treatment. Can treatment duration and  need  for  alternative  therapies  be  guided  by  the results  of  the  minimal  residual  disease  assessment? This  question  will  have  to  be  answered  prospectively through well-designed response-adapted clinical trials. 最后，重链/轻链比值的使用可能对确定最小残留病灶阴性状态具有重要作用。 细胞检测阴性、正电子发射计算机断层扫描阴性以及重链/轻链比值正常，这三者的结合可能代表了一种综合终点，它反映了肿瘤细胞从所有部位的根除以及正常浆细胞群恢复到目前可用的检测水平。这方面需要在前瞻性临床试验和大型回顾性数据集中进一步研究。开发基于血液的检测方法，无论是检测罕见循环细胞还是循环肿瘤 DNA，都将是理想的选择，目前的工作重点应放在开发这些方法上。 随着灵敏度更高的基于血流的检测方法越来越普遍，我们预计严格的完全反应标准将越来越少地被使用，最终可能会被取消。这一因素尤为重要，因为作为严格的完全反应标准的一部分，sFLC 正常化的贡献受到了法语髓样组织间小组数据的质疑。 如果严格完全反应比完全反应更有用，主要是因为用灵敏度较低的方法无法检测到浆细胞，那么使用最小残留病方法将使这一标准过时。另一个正在积极研究的领域是用 sFLC 测量替代 24 小时尿液测量。虽然这种替代将大大减轻患者和医生的负担，但目前还没有明确的数据支持这种改变。118-120 这种方法提出的最重要问题是最小残留病结果对治疗决策的影响。最小残留病灶的评估结果能否指导治疗持续时间和替代疗法的需求？这个问题需要通过精心设计的反应适应性临床试验来回答。

Search strategy and selection criteria 搜索策略和选择标准

We searched PubMed for articles published in English between Jan 1, 1980,and June 30, 2014, that contained the term “myeloma” and any one of the following terms: “response” or “minimal residual disease” or “imaging” or “bone marrow” or “monoclonal protein”. We also reviewed recent reviews on multiple myeloma. Members of the International Myeloma Working Group were then asked to    identify any appropriate citation that was of interest but not detected by the search strategy. 我们在 PubMed 上搜索了 1980 年 1 月 1 日至 2014 年 6 月 30 日期间发表的英文文章，其中包含 " myeloma "和以下任一术语：" response "或 " minimal residual disease "或 " imaging "或 " bone marrow "或 " monoclonal protein "。我们还查阅了近期有关多发性骨髓瘤的综述。随后，国际骨髓瘤工作组成员被要求找出任何感兴趣但搜索策略未发现的适当引文。

 Panel: Practical considerations for application of IMWG consensus criteria  小组讨论：应用 IMWG 共识标准的实际考虑因素 

•    If partial or minimal response rate is an endpoint, patients must have measurable disease at baseline, within the window defined by the study protocol; if multiple measurements are available, the measurement closest to cycle 1, day 1 will be used as baseline - 如果以部分或最小反应率作为终点，患者必须在研究方案规定的时间窗内，在基线时有可测量的疾病；如果有多个测量值，则以最接近第 1 周期第 1 天的测量值作为基线

•    If patients do not have measurable disease at baseline they can only be assessed for at least a complete response or progressive disease - 如果患者在基线时没有可测量的疾病，则只能对其进行至少完全应答或疾病进展的评估

•    Measurable disease is defined as - 可测量的疾病是指

1. Serum M-protein ≥1 g/dL 1.血清 M 蛋白≥1 克/分升

2. Urine M-protein ≥200 mg/24 h 2.尿液中的 M 蛋白≥200 毫克/24 小时

3. Serum FLCassay: involved FLC level ≥10 mg/dL provided serum FLC ratio is abnormal 3.血清 FLC 检测：如果血清 FLC 比值异常，则涉及 FLC 水平≥10 mg/dL

•    Missing serum and/or urine electrophoresis during disease follow-up remains a significant problem. In general, the following considerations will allow a more uniform assessment: - 在疾病随访期间，血清和/或尿液电泳的缺失仍是一个重要问题。一般来说，以下几点可使评估更加统一：

1. In the context of a clinical trial, missing serum or urine electrophoresis, or both, can only be accepted at the    discretion of an independent review committee 1.在临床试验中，血清或尿液电泳结果缺失，或两者同时缺失，只能由独立审查委员会决定是否接受。

2.  If the immunofixation of the serum or urine is negative  at baseline,any lack of follow-up testing of the serum or urine can be accepted at the discretion of the independent review committee 2.如果基线时血清或尿液的免疫反应为阴性，独立审查委员会可酌情接受任何未进行后续血清或尿液检测的情况。

3. Parameters that are considered measurable at baseline (serum and urine, FLC if both serum and urine are not  measurable) should be performed at each assessment 3.每次评估都应检测基线时可测量的参数（血清和尿液，若血清和尿液均不可测量，则检测 FLC）。

4. Urine M-protein is not needed to document partial response or minor response if baseline urine M-protein was not measurable; however, it is still required for complete response and very good partial response 4.如果基线尿液 M 蛋白无法测量，则不需要尿液 M 蛋白来记录部分应答或轻微应答；但是，完全应答和非常好的部分应答仍需要尿液 M 蛋白。

•    A plasmacytoma that has been radiated is not suitable for response assessment; however, it must be monitored to   assess for progressive disease - 已接受放射治疗的浆细胞瘤不适合进行反应评估；但必须对其进行监测，以评估疾病的进展情况

•    A baseline bone marrow examination must always be attempted; if the patient declines or if the sampling is unsuccessful this must be documented; when bone marrow plasma-cell infiltration is assessed by both bone marrow aspirate and by bone marrow biopsy, the highest value of bone marrow plasma-cell infiltration should be used - 必须始终尝试进行骨髓基线检查；如果患者拒绝或取样不成功，必须记录在案；如果骨髓浆细胞吸出术和骨髓活检均可评估骨髓浆细胞吸出量，则应使用骨髓浆细胞吸出量的最高值。

•    For patients achieving very good partial response by other    criteria, a soft tissue plasmacytoma must decrease by more   than 90% in the sum of the maximal perpendicular diameter (SPD) compared with baseline - 对于通过其他标准获得非常好的部分反应的患者，软组织浆细胞瘤的最大垂直直径（SPD）之和必须比基线下降 90% 以上

•    Single discrepant results can be ignored at the discretion of an independent review committee - 独立审查委员会可酌情忽略单个不一致的结果

•    For IgA and IgD myelomas, quantitative immunoglobulin measurements are preferred for disease assessments; the same percentage changes applies as for serum M-spike - 对于 IgA 和 IgD 骨髓瘤，疾病评估首选定量免疫球蛋白测量；百分比变化与血清 M-尖峰蛋白的变化相同。

•    Serum FLC levels should only be used for response assessment when both the serum and urine M-component levels are deemed not measurable - 只有当血清和尿液中的 M 组份水平均被认为无法测量时，血清 FLC 水平才可用于反应评估

•    Documentation of response requires two consecutive readings of the applicable disease parameter (serum M-protein, urine M-protein, or serum FLC),performed at  anytime (no minimum interval is required, it can be done the same day); however, to confirm response or progressive disease, two discrete samples are required; testing cannot be based upon the splitting of a single   sample - 反应的记录要求在任何时间连续读取两个适用的疾病参数（血清 M 蛋白、尿液 M 蛋白或血清 FLC）（不要求最短间隔时间，可在同一天进行）；但是，要确定反应或疾病进展，需要两个不连续的样本；检测不能基于单个样本的拆分。

•    Whenever more than one parameter is used to assess response, the overall assigned level of response is determined by the lower or lowest level of response - 在使用一个以上参数评估响应时，响应的总体分配水平由较低或最低的响应水平决定

•    Patients should be categorised as stable disease until they meet criteria for any response category or have progressive disease - 在达到任何反应类别的标准或疾病进展之前，患者应被归类为病情稳定患者

•    Patients will continue in the last confirmed response category until there is confirmation of progression or improvement to a higher response status; patients cannot move to a lower response category - 患者将继续留在最后确定的反应类别中，直到确定病情进展或改善到更高的反应状态；患者不得转入较低的反应类别

•    If alternate therapy is started before confirming progressive disease any additional testing during subsequent therapy    can be used to confirm progressive disease - 如果在确定疾病进展之前就开始了替代疗法，那么在随后的治疗过程中进行的任何额外检测都可用于确定疾病进展。

•    The lowest confirmed value before suspected progression will be used as baseline for calculation of progression; if a  serum and/or urine spike is considered too low to quantitate, this value can be assigned as zero as a baseline for documentation of subsequent progressive disease - 疑似病情进展前的最低确证值将作为计算病情进展的基线；如果血清和/或尿液中的峰值被认为太低而无法定量，则可将该值定为零，作为记录随后病情进展的基线。

•    Any soft tissue plasmacytoma documented at baseline must undergo serial monitoring; otherwise, the patient is classified as inevaluable - 基线时记录的任何软组织浆细胞瘤都必须接受连续监测；否则，患者将被归类为无价值病例

•    Patients will be considered to have progressive disease if they meet the criteria for progression by a variable that was not considered measurable at baseline; however, for patients who had a measurable serum or urine M-spike at baseline, progression cannot bedefined by increases in     serum FLC alone - 然而，对于基线时血清或尿液中存在可测量的 M-尖峰蛋白的患者，不能仅凭血清 FLC 的升高来判定病情进展。

•    Inpatients with two monoclonal protein bands at the start of therapy, the sum of the two spikes should be used for monitoring of disease - 在开始治疗时有两个单克隆蛋白条带的住院病人，应使用两个峰值的总和来监测病情

•    Careful attention should be given to new positive immunofixation results appearing inpatients who have achieved a complete response, when the isotype is different, it probably represents oligoclonal immune reconstitution and should not be confused with relapse; these bands typically disappear over time - 对于已获得完全应答的患者出现的新的免疫检测阳性结果，如果同种型不同，则可能代表少克隆免疫重建，不应与复发混淆；这些条带通常会随着时间的推移而消失。

FLC=游离轻链。IMWG=国际骨髓瘤工作组