ORIGINAL ARTICLE

A multiparameter flow cytometry immunophenotypic algorithmfor the identification of newly diagnosed symptomatic myelomawith an MGUS-like signature and long-term disease controlThis article has been corrected since Advance Online Publication and a corrigendum is also printed in this issue

B Paiva1,2, M-B Vıdriales1,2, L Rosinol3, J Martınez-Lopez4, M-V Mateos1,2, EM Ocio1,2, M-A Montalban4, L Cordon5, NC Gutierrez1,2,L Corchete1,2, A Oriol6, M-J Terol7, M-A Echeveste8, R De Paz9, F De Arriba10, L Palomera11, J de la Rubia5, J Dıaz-Mediavilla12,M Granell13, A Gorosquieta14, A Alegre15, A Orfao2,16, J-J Lahuerta4, J Blade3 and JF San Miguel1,2 on behalf of the GEM(Grupo Espanol de MM)/PETHEMA (Programa para el Estudio de la Terapeutica en Hemopatıas Malignas) cooperative study group

Achieving complete remission (CR) in multiple myeloma (MM) translates into extended survival, but two subgroups of patients falloutside this paradigm: cases with unsustained CR, and patients that do not achieve CR but return into a monoclonal gammopathyof undetermined significance (MGUS)-like status with long-term survival. Here, we describe a novel automated flow cytometricclassification focused on the analysis of the plasma-cell compartment to identify among newly diagnosed symptomatic MMpatients (N¼ 698) cases with a baseline MGUS-like profile, by comparing them to MGUS (N¼ 497) patients and validating theclassification model in 114 smoldering MM patients. Overall, 59 symptomatic MM patients (8%) showed an MGUS-like profile.Despite achieving similar CR rates after high-dose therapy/autologous stem cell transplantation vs other MM patients, MGUS-likecases had unprecedented longer time-to-progression (TTP) and overall survival (OS; B60% at 10 years; Po0.001). Importantly,MGUS-like MM patients failing to achieve CR showed similar TTP (P¼ 0.81) and OS (P¼ 0.24) vs cases attaining CR. This automatedclassification also identified MGUS patients with shorter TTP (P¼ 0.001, hazard ratio: 5.53) and ultra-high-risk smoldering MM(median TTP, 15 months). In summary, we have developed a biomarker that identifies a subset of symptomatic MM patients with anoccult MGUS-like signature and an excellent outcome, independently of the depth of response.

Leukemia (2013) 27, 2056–2061; doi:10.1038/leu.2013.166

Keywords: myeloma; plasma cells; flow cytometry; immunophenotyping; survival; Euroflow

INTRODUCTIONProlongation of patients’ survival after the introduction of high-dose therapy (HDT) and novel agents into the up-front treatmentof multiple myeloma (MM) is now unquestionable.1 However, theincrement in response rates,2,3 life expectancy4 and differenttreatment alternatives5–8 has also raised new questions, eitherrelated to the importance of the depth of remission achievedor specific treatment decisions (for example, timing fortransplantation and duration of maintenance).9 The importanceof achieving complete remission (CR) after front-line therapy toextend progression-free survival is almost consensual,10,11 withtwo potential groups of patients justifying the lack of globalagreement: those that show unsustained CR after treatment andare associated to a dismal outcome, and those that despite notachieving CR, experience long-term disease control.12 Recentobservations have shed some light about the prospectiveidentification of patients at risk of showing unsustained CR(those cases with high-risk cytogenetics and persistent minimal

residual disease),13 whereas consistent data on patients failing toachieve CR but with an indolent disease course remains mostlyrestricted to a single study on the total therapy programs, whichshows that cases with a monoclonal gammopathy of undeter-mined significance (MGUS)-like signature at baseline by geneexpression profiling are associated with 410-year survival ratesafter HDT/autologous stem cell transplantation (ASCT) despitedisplaying a lower incidence of CR.14 The prospective identi-fication of this particular subgroup of MM patients may be ofutmost clinical relevance in order to avoid overtreatment inpatients that otherwise would have been considered as sub-optimal responders; therefore, investigation of new biomarkersand easy to perform assays for routine diagnostic laboratoriescould potentially contribute to identify such MM patients.

Here, we investigate the potential utility of a multiparameterflow cytometry (MFC) immunophenotyping-computerized algo-rithm based on the simultaneous assessment of the tumor burdenand the degree of clonality of the bone marrow (BM) plasma-cell

1Hospital Universitario de Salamanca, Salamanca, Spain; 2IBSAL, IBMCC (USAL-CSIC), Salamanca, Spain; 3Hospital Clınic, IDIBAPS, Barcelona, Spain; 4Hospital 12 de Octubre,Madrid, Spain; 5Hospital Universitario de la Fe, Universidad Catolica de Valencia San Vicente Martir, Valencia, Spain; 6Hospital Universitari Germans Trias i Pujol, Badalona, Spain;7Hospital Clınico de Valencia, Valencia, Spain; 8Hospital de Donostia, San Sebastian, Spain; 9Hospital Universitario La Paz, Madrid, Spain; 10Hospital Morales Meseguer, Murcia,Spain; 11Hospital Lozano Blesa, Zaragoza, Spain; 12Hospital Clınico San Carlos, Madrid, Spain; 13Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; 14Hospital UniversitarioNavarra, Pamplona, Spain; 15Hospital Universitario La Princesa, Madrid, Spain and 16Servicio General de Citometrıa and Department of Medicine, Universidad de Salamanca,Salamanca, Spain. Correspondence: Professor JFS Miguel, Hospital Universitario de Salamanca, Paseo de San Vicente 58–182 37007 Salamanca, Spain.E-mail: sanmigiz@usal.esReceived 7 May 2013; revised 27 May 2013; accepted 30 May 2013; accepted article preview online 7 June 2013; advance online pubication, 25 June 2013

Leukemia (2013) 27, 2056–2061& 2013 Macmillan Publishers Limited All rights reserved 0887-6924/13

www.nature.com/leu

(PC) compartment to identify symptomatic MM patients with anMGUS-like profile at baseline who are associated with long-termdisease control, irrespective of the depth of remission achievedafter treatment. In addition, we have investigated the value of thisalgorithm to predict for the risk of malignant transformation ofMGUS and smoldering MM.

PATIENTS AND METHODSPatients and treatmentThe study involved 698 newly diagnosed MM patients included in twoconsecutive PETHEMA/GEM trials: GEM2000 (VBMCP/VBAD followed byHDT/ASCT and 2 years of maintenance with interferon plus prednisone;N¼ 486)15 and GEM2005o65y (randomized induction with the samechemotherapy plus bortezomib in the last two cycles or thalidomide/dexamethasone or bortezomib/thalidomide/dexamethasone followed byHDT/ASCT, and 3 years of maintenance with interferon-a2b or thalidomideor thalidomide/bortezomib; N¼ 212).16 Median follow-up was 71 months(range: 4–153 months). To investigate for an MGUS-like profile amongpatients with symptomatic MM, we have compared them with 497 MGUSpatients. In addition, a series of 114 smoldering MM patients wasused as a validation set. All cases (N¼ 1309) were diagnosed accordingto the International Myeloma Working Group (IMWG) criteria,17 and allsamples were collected after informed consent was given; the study wasapproved by the local ethical committees.

MFC immunophenotypic studiesErythrocyte-lysed whole BM samples were immunophenotyped using afour-color combination of the following fluorochrome-conjugated—fluorescein isothiocyanate, phycoerythrin, peridinin chlorophyll protein-cyanin 5.5 (PerCPCy5.5) and allophycocyanin—monoclonal antibodies:CD38 CD56 CD19 CD45. PCs were initially identified on the basis of strongCD38 expression and intermediate side scatter signals. Clonal PCs werediscriminated from normal PCs by aberrant phenotypic expressions,typically consisting of simultaneous downregulation of CD19 and CD45,with or without overexpression of CD56 (in around 2/3 of the cases).

For the remaining patients in whom CD45 was positively expressed, lack ofCD19 and/or bright CD56 staining (equal or higher than that of naturalkiller cells) allowed identification of clonal PCs.18 Data acquisition wasperformed in FACSCalibur and FACSCantoII flow cytometers (BectonDickinson Biosciences, San Jose, CA, USA) using the FACSDiva 6.1 software(Becton Dickinson Biosciences), and a two-step acquisition procedure.In the first step, information about 5� 104 events correspondingto the whole-sample cellularity was stored; in the second step, dataabout CD38þgated events was selectively stored, for a minimum of106 leukocytes/tube. Data analysis was performed using the Infinicytsoftware (Cytognos SL, Salamanca, Spain).

Generation of a computerized model based on MFCimmunophenotypic analysisThe relative frequency of BM PCs plus the percentage of clonal andnormal PCs within the whole BM PC compartment was determinedfor each patient. Subsequent principal component analysis (PCA)using these three variables was then performed and shown by theautomatic population separator (APS; principal component 1 vs principalcomponent 2) graphical representation of the Infinicyt software.19 In thisAPS view, every patient becomes represented by a dot, and the first (x axis)and second (y axis) principal components are used to produce a bi-dimensional representation of patient profiles; each principal componentis represented by a linear combination of the three parameters describedabove.20 On the basis of this APS representation, MGUS and symptomaticMM patients were simultaneously grouped with each patient representedby a single dot and groups defined by 1.5 standard deviation (s.d.) medianborders (Figure 1).

Statistical analysesThe Mann–Whitney U-test was used to estimate the statistical significanceof differences observed between groups. MGUS and smoldering MMpatients who had not progressed were censored on the last date theywere known to be alive and without progression. Overall survival (OS) wasmeasured from the start of chemotherapy to the date of death or last visit.Time-to-progression (TTP) and OS curves were plotted by the Kaplan–Meier method, and the log-rank test was used to estimate the statisticalsignificance of differences observed between curves. Cox stepwiseregression (backward elimination) was used for multivariate analysis,retaining those variables with a statistically significant predictivevalue (Pp0.05) in the predictive model. For all statistical analyses, theSPSS software (version 15.0; SPSS Inc., Chicago, IL, USA) was used.

RESULTSIdentification of newly diagnosed symptomatic MM patients withan MGUS-like profileA total of 59 patients with symptomatic MM (8%) were plotted inthe APS graphical representation (Figure 1) outside of theirreference MM group (N¼ 698) and inside the MGUS cluster(N¼ 497), such cases were being thereby termed as MGUS-likefrom now on. All 59 patients had at least one CRAB feature(7% with increased Calcium, 3% with Renal insufficiency, 13% withAnemia and 92% with detectable Bone lesions). MGUS-like MMpatients as compared with the other cases (Table 1) werecharacterized by a favorable clinical presentation with a higherfrequency of International Staging System stage I or II (89% vs75%; Po0.001), greater baseline hemoglobin levels (medianof 123 vs 104 g/l, respectively; Po0.001), as well as lowertumor burden as assessed either by serum paraprotein levels(2.2 vs 4.0 g/dl; Po0.001), PCs by conventional morphology(10% vs 41% PCs; Po0.001) or by immunophenotyping (0.6% vs12%, Po0.001). In addition, normal PCs within the total BM PCcompartment were preserved (32% vs 0.3% of all BM PCs,Po0.001), and clonal PCs from MGUS-like MM patients were morequiescent (0.7% vs 1.2%; P¼ 0.02). Age at presentation was similarbetween the two groups (median of 58 years). When comparingthe phenotype of clonal PCs in the two groups of patients, only atrend toward lower frequency of CD81þ clonal PCs in MGUS-likepatients as compared with the other cases (20% vs 51%; P¼ 0.06)

Figure 1. PCA classification model based on MFC immunopheno-typic analysis of reference MGUS and symptomatic MM patients.The relative frequency of BM PCs plus the percentage of clonal andnormal PCs within the BM PC compartment was determined foreach individual patient. In addition, subsequent PCA of MGUS andsymptomatic MM patients based on these variables was performed,and results represented in a principal component 1 (x axis) vsprincipal component 2 (y axis) APS graphical representation(Infinicyt software). On the basis of this APS representation,individual MGUS (green dots) and symptomatic MM patients (reddots) are plotted as two potentially overlapping clusters by 1.5 s.d.curves. A total of 59 symptomatic MM patients (8%) were classifiedoutside of their reference group (outside 1.5 s.d. curves) and insidethe MGUS cluster, and they were thereby termed as MGUS-likesymptomatic MM patients.

Flow cytometric MGUS-like profile in myelomaB Paiva et al

2057

& 2013 Macmillan Publishers Limited Leukemia (2013) 2056 – 2061

was observed (Table 2). No significant differences were noted forPC DNA ploidy status. Regarding cytogenetic abnormalities,del(13q) was uncommon among MGUS-like MM patients (6% vs45%; P¼ 0.002), and there was also a trend for lower incidence ofIGH translocations (18% vs 40%; P¼ 0.06), particularly for thoseassociated with high-risk disease (Table 2).

Outcome of newly diagnosed symptomatic MM patients with anMGUS-like profileWe first investigated whether symptomatic MM patients with anMGUS-like profile at baseline were more likely to attain CR afterHDT/ASCT as compared with the rest of the cases, but nosignificant differences were noted (50% vs 41%, respectively;P¼ 0.21). However, MGUS-like MM patients did show a signifi-cantly superior TTP (median not reached (NR) vs 44 months,Po0.001; hazard ratio: 3.27; Figure 2a) and OS (median NR vs 67months, Po0.001, hazard ratio: 2.54; Figure 2b). Notably, the 10-year estimate for TTP and OS in MGUS-like patients were 59% and65%, respectively (Po0.001). As the superior outcome of MGUS-like MM cases was not foreseen by the rates of CR attained after

HDT/ASCT, we then investigated if this model maintained itsprognostic value among patients failing to achieve CR. Accord-ingly, among symptomatic MM cases who did NR CR, patientsclassified by the computerized algorithm as MGUS like at baselinealso showed a superior TTP (median NR vs 38 months, Po0.001)and OS (median NR vs 54 months, P¼ 0.003) than the othersymptomatic MM cases. Moreover, when focusing exclusively onMGUS-like symptomatic MM patients, no significant differencesfor median TTP and OS were noted between cases in CR vs thosethat did not attain CR after HDT/ASCT (Figure 3). By contrast, ascould be expected, achieving CR among non-MGUS-like MMpatients predicted for significantly (Po0.001) longer TTP and OS. Itshould be noted that around one-fourth (28%) of all MGUS-likeMM patients were progression-free survival at 10 years (vs 4% ofthe other symptomatic MM cases; Po0.001).

Prediction of malignant transformation of MGUS patients by theMFC immunophenotypic-computerized algorithmWhen we applied the same MFC immunophenotypic classificationmodel to MGUS cases, a few of these (7/497; 1%) were plottedoutside of their reference group, clustering together with typicalsymptomatic MM (Figure 4). These patients had a markedimbalance between clonal and normal PCs (99%–1%) similar tothat found in symptomatic MM, and/or a relative high frequencyof BM PCs by MFC (9 to 13%) although all had o10% bymorphology; interestingly, three of these MGUS cases (43%) didprogressed into symptomatic MM with a median TTP for thispatient population of 85 months vs 234 months for the standardMGUS cases (P¼ 0.001, hazard ratio: 5.53; Figure 4).

Validation of the MFC immunophenotypic-computerizedalgorithm in smoldering MMWe finally assessed if the MFC immunophenotypic-computerizedclassification model generated herein could maintain its signifi-cance in a different series of patients. For this purpose, we havefaced 114 newly diagnosed smoldering MM patients, andclassified them against the MGUS and symptomatic MM referencegroups. Our results showed the expected heterogeneity thatcharacterizes smoldering MM patients, with approximately half ofthe cases (51%) being plotted in the overlapping area betweenMGUS and symptomatic MM, 41 patients (36%) being classified asMGUS-like smoldering MM and the remaining 15 cases (13%)being assigned to the symptomatic MM reference group(Figure 5). Smoldering MM patient stratified according to thismodel had different baseline clinical features (SupplementaryTable 1) and, importantly, it translated into markedly different TTPinto active disease: median NR and 5% risk of progression at 2years for MGUS-like smoldering MM cases; median of 108 monthsand 28% risk of progression for intermediate patients; and, finally,median TTP of 15 months and 53% risk of progression at 2 yearsfor symptomatic MM-like smoldering MM patients.

DISCUSSIONThere is increasing body of evidence that MM should no longer beconsidered as a single entity,21–25 and similar to what occurs inother hematological malignancies, such as acute leukemias andnon-Hodgkin lymphomas, different subtypes need to berecognized in order to define tailored therapies. Currently,(cyto)genetics is the most relevant tool used in MM todiscriminate different risk categories, although this has not yetbeen translated into treatment adaptation.26–28 The secondrelevant prognostic biomarker is response to therapy, CR beingassociated with longer survival; although this may guideconsolidation and maintenance therapy decisions, there aresubgroups of patients that do fall outside of the CR paradigm:those showing unsustained CR12,13 and those that do not achieveCR because they may have returned into an MGUS stage. How to

Table 1. Demographics and baseline clinical characteristics ofsymptomatic MM patients with an MGUS-like profile (N¼ 59) bycomputerized MFC immunophenotyping vs other symptomatic(non-MGUS-like) MM patients

Symptomatic MM MGUSlike

Non MGUSlike

P-value

(n¼ 59) (n¼ 639)

Male, % 68 55 0.08Age, year (median) 58 58 0.3IgG/ IgA/light chain, % 63/17/19 55/25/18 0.7ISS stage I/ II/ III,% 57/32/11 32/43/25 o0.001Hemoglobin, g/l (median) 123 104 o0.001b2 microglobulin, mg/l (median) 2.3 3.4 o0.001Albumin, g/dl (median) 3.9 3.5 0.005M-component, g/dl (median) 2.2 4.0 o0.001PCs by CM, % (median) 10 41 o0.001PCs by MFC, % (median) 0.6 12 o0.001aNormal PCs by MFC, % (median) 32 0.3 o0.001

Abbreviations: CM, conventional morpology; Ig, immunoglobulin; ISS,International Staging System; PCs, plasma cells; MFC, multiparameter flowcytometry; MM, multiple myeloma. aPercentage of normal plasma cellswithin the bone marrow plasma cell compartment.

Table 2. Biological characteristics of clonal PCs in MM patients with anMGUS-like profile by computerized MFC immunophenotyping vsother symptomatic (non-MGUS-like) MM patients

Symptomatic MM MGUS-like

Non MGUSlike

P-value

aCD81þ expression on clonal PCs byMFC, %

20 51 0.06

Hyperdiploid DNA content, % 58 51 0.4Plasma cells in S-phase, % 0.7 1.2 0.02

Cytogenetics, %IgH translocations 18 40 0.06t(4;14) 0 11 0.1t(11;14) 6 15 0.3t(14;16) 0 4 0.4other t(IgH) 12 10 0.8

del(13q) 6 45 0.002del(17p) 6 7 0.9bHigh-risk cytogenetics 6 20 0.2

Abbreviations: Ig, immunoglobulin; MFC, multiparameter flow cytometry;MM, multiple myeloma; PCs, plasma cells. aP-value 40.1 for other antigensunder study (for example, CD19, CD27, CD28, CD33, CD38, CD45, CD56,CD81, CD117 and b2 microglubulin). bHigh-risk cytogenetics defined bythe presence of t(4;14), t(14;16) and/or del(17p).

Flow cytometric MGUS-like profile in myelomaB Paiva et al

2058

Leukemia (2013) 2056 – 2061 & 2013 Macmillan Publishers Limited

discriminate this latter subgroup of MM patients that would beassociated with stable disease and prolonged survival, fromresistant patients with suboptimal responses and the likelihood ofprogressive disease remains a challenge. This task should notbe taken with lightness because these patients should bediscriminated from true non-responders, and therapeuticdecisions regarding further intensification and consolidation/maintenance strategies should be taken accordingly. Here, hehave strengthen MFC immunophenotyping using novel softwareanalysis tools to develop an automated classification modelcapable of identifying newly diagnosed symptomatic MM patientswith a baseline MGUS-like profile. This small subset (8% of allcases) shows an unprecedented TTP of 59% at 10 years, and itsprognosis is not dependent on the depth of response achieved(CR vs no CR).

Long-term follow-up studies in the setting of HDT/ASCT arenow showing that only a small fraction of MM patients (6–18%)remain X10-year-relapse free, and these are now considered asbeing operationally cured.29–31 Some biomarkers have beenproposed to identify patients with long-term survival, namely agene expression profiling signature of MGUS (detected in B30%of patients) and the CD2 molecular subtype.14 There is growingbody of evidence showing that the risk of progression of MGUS,and particularly smoldering MM patients, into symptomatic MMcan be predicted by using markers that evaluate the extent oftumor burden (for example, M-protein, marrow plasmacytosis,focal lesions by MRI or circulating PCs32–38), or markers related tothe level of clonal excess as reflected by the balance betweennormal and clonal PCs (for example, serum-free light or heavy-light chains, immune paresis and clonal excess by MFC).37,39–42

Figure 2. TTP (a) and OS (b) of symptomatic MM patients classified according to their MFC immunophenotypic profile defined by thePCA-based classification algorithm developed (Figure 1).

Figure 3. TTP (a) and OS (b) of MGUS-like symptomatic MM patients classified according to the depth of response to therapy.

Figure 4. TTP into active myeloma of MGUS patients (b) classified outside their reference group and inside the cluster correspondingto of symptomatic MM patients (N¼ 7; 1%) according to the PCA-based classification algorithm generated with MFC immunophenotypicvariables (a).

Flow cytometric MGUS-like profile in myelomaB Paiva et al

2059

& 2013 Macmillan Publishers Limited Leukemia (2013) 2056 – 2061

Here, we postulated that the same principle could be inverselyapplied to symptomatic MM patients, in order to potentiallyidentify a subgroup of MM patients with an MGUS-like profile.Based on PCA analysis of three easily accessible immuno-phenotypic variables (the relative frequency of BM PCs plus thepercentage of clonal and normal PCs within the whole-BM PCcompartment), here we propose a classification algorithmconstructed through the analysis of a large number of MGUSand symptomatic MM patients, its results being graphicallyvisualized with currently available MFC software (for example,APS). Based on the proposed algorithm, we identified 8% ofsymptomatic MM patients clustering outside of their referencegroup and together with typical MGUS cases. These patientspresented with favorable baseline clinical features and lessproliferative disease (which may be in accordance with a lowerfrequency of CD81þ clonal PCs43). In addition, their clinical(symptomatic) recognition appears to occur before theaccumulation of early cytogenetic abnormalities such as del(13q)and IGH translocations. CR rates after HDT/ASCT were similaramong the two groups, but TTP and OS were markedly superioramong symptomatic MM patients with an MGUS-like profile;moreover, the outcome of these patients was not compromisedby failure of achieving CR.

From a translational point of view, the identification ofbiomarkers that may help to guide therapeutic decisions is likelyto become more and more important with the availability of newtreatment protocols, particularly regarding intensity and durationof therapy. Ideally, such biomarkers should be accurate, easy tostandardized and provide results in a timely and cost-effectiveway. In the present study, we adopted new analytic toolsdeveloped by the EuroFlow consortium19 to build a referencedata set of MGUS and MM patients based on MFC characterizationof the PC compartment. Because it relies in more than 1000patients and three immunophenotypic variables, outliers canpotentially be more accurately identified. Accordingly, amongMGUS patients only seven cases were ‘misclassified’ as MM but,interestingly, three have now evolved into symptomatic diseasewith an overall estimated risk at 10 years of 67%. This estimatedrate of transformation is superior to that found in our ownprevious observations (56% for MGUS patients with 495% clonalPCs among total BM PCs)42 and those found with PC counts bymorphology (45%)34 or serological markers such as the M-protein(415 g/l),32 abnormal sFLCs,40 or HLC-pair suppression.41 Likewise,we also identified among smoldering MM patients a group ofapproximately one-third of cases showing a clinical course similarto MGUS (those SMM patients with an MGUS-like profile) and,conversely, a smaller group of 10% smoldering MM cases with asymptomatic MM-like profile (median frequency of BM PCs of

11%, and a balance between clonal and normal PCs of 99%–1%)and median TTP of only 15 months. These patients could probablybe considered as having ultra-high risk of progression or ‘early MM’,44

and, again, the computerized classification model here generatedsupersedes our own previous observations (median TTP of34 months for SMM patients with 495% clonal PCs among totalBM PCs), and matches recent results obtained by morphological PCcounts above 60%36,45 and highly abnormal sFLC ratios.36,37

In summary, we have developed a MFC immunophenotyping-based classification algorithm capable of identifying newly diagnosedsymptomatic MM patients with a baseline MGUS-like profile,which show long-term survival independent of CR achievement.The prospective identification of this signature may contribute todiscriminate a suboptimal response that requires additionaltreatment from a residual ‘MGUS-like component’ that mayremain stable without further treatment. In addition, the modelalso contributes to the identification of MGUS and smoldering MMpatients at high risk of malignant transformation. Importantly, thereference data set and classification algorithm here developed canbe equally built or shared across different myeloma centers.

CONFLICT OF INTERESTJM-L, EMO, A Oriol, FdA, LP, JdlR and J-JL have received honoraria from Celgene andJanssen. BP has received honoraria from Millennium, Janssen, Celgene and theBinding Site. M-VM has served on the speaker’s bureau for Millennium, Celgene andJanssen. LR has received honoraria and has served on the advisory board for Janssenand Celgene, JB has received honoraria and has served on the advisory board forJanssen and Celgene, as well as grant support from Celgene and Janssen. JFSM hasserved on the speaker’s bureau and on the advisory board for, and has receivedhonoraria from, Millennium, Janssen and Celgene. All other authors declare noconflict of interest.

ACKNOWLEDGEMENTSThis work was supported by the Cooperative Research Thematic Network (RTICC)grants RD12/0036/0071, RD12/0036/0058, RD12/0036/0046, RD12/0036/0036 andRD12/0036/0048, of the Red de Cancer (Cancer Network of Excellence), Instituto deSalud Carlos III, Spain, Instituto de Salud Carlos III/Subdireccion General de InvestigacionSanitaria (FIS: PI060339; 06/1354; 02/0905; 01/0089/01-02; PS09/01897/01370; G03/136),Consejerıa de Sanidad, Junta de Castilla y Leon, Valladolid, Spain (557/A/10) andAsociacion Espanola Contra el Cancer (AECC; GCB120981SAN), Spain.

AUTHOR CONTRIBUTIONSJFSM, J-JL, JB and A Orfao conceived the idea, and together with M-VM and LRdesigned the study protocol; BP, M-BV, M-AM, and LC analyzed the flowcytometry data; NCG analyzed the FISH data; LR, JM-L, M-VM, EMO, AO, M-JT,M-AE, RdP, FdA, LP, JdlR, JD-M, MG, AG, AA, J-JL, JB and JFSM contributed with

Figure 5. TTP into active myeloma of smoldering SMM patients classified according to the PCA model based on MFC immunophenotypic datahere proposed. Patients plotted in the overlapping area between MGUS and symptomatic MM (N¼ 58; 51%), classified as MGUS-like (N¼ 41;36%), or assigned to the symptomatic MM reference group (N¼ 15; 13%) (a) showed markedly different TTP (b).

Flow cytometric MGUS-like profile in myelomaB Paiva et al

2060

Leukemia (2013) 2056 – 2061 & 2013 Macmillan Publishers Limited

provision of study material or patients; BP and JFSM analyzed and interpretedthe data; BP and LC performed the statistical analysis; and BP and JFSM wrotethe manuscript. All authors reviewed and approved the manuscript.

REFERENCES1 San-Miguel JF, Mateos MV. Can multiple myeloma become a curable disease?

Haematologica 2011; 96: 1246–1248.2 Stewart AK, Richardson PG, San-Miguel JF. How I treat multiple myeloma in

younger patients. Blood 2009; 114: 5436–5443.3 Leleu X, Fouquet G, Hebraud B, Roussel M, Caillot D, Chretien ML et al.

Consolidation with VTd significantly improves the complete remission rate andtime to progression following VTd induction and single autologous stem celltransplantation in multiple myeloma. Leukemia 2013; e-pub ahead of print6 April 2013.

4 Usmani SZ, Crowley J, Hoering A, Mitchell A, Waheed S, Nair B et al. Improvementin long-term outcomes with successive Total Therapy trials for multiple myeloma:are patients now being cured? Leukemia 2013; 27: 226–232.

5 Palumbo A, Anderson K. Multiple myeloma. New Engl J Med 2011; 364: 1046–1060.6 Palumbo A, Cavallo F. Have drug combinations supplanted stem cell transplan-

tation in myeloma? Blood 2012; 120: 4692–4698.7 San-Miguel JF. Consolidation therapy in myeloma: a consolidated approach?

Blood 2012; 120: 2–3.8 Rajkumar SV. Multiple myeloma: 2012 update on diagnosis, risk-stratification, and

management. Am J Hematol 2012; 87: 78–88.9 Badros AZ. Lenalidomide in myeloma--a high-maintenance friend. New Engl J Med

2012; 366: 1836–1838.10 Chanan-Khan AA, Giralt S. Importance of achieving a complete response in multiple

myeloma, and the impact of novel agents. J Clin Oncol 2010; 28: 2612–2624.11 Harousseau JL, Attal M, Avet-Loiseau H. The role of complete response in multiple

myeloma. Blood 2009; 114: 3139–3146.12 Barlogie B, Anaissie E, Haessler J, van Rhee F, Pineda-Roman M, Hollmig K et al.

Complete remission sustained 3 years from treatment initiation is a powerfulsurrogate for extended survival in multiple myeloma. Cancer 2008; 113: 355–359.

13 Paiva B, Gutierrez NC, Rosinol L, Vidriales MB, Montalban MA,Martinez-Lopez J et al. High-risk cytogenetics and persistent minimal residualdisease by multiparameter flow cytometry predict unsustained completeresponse after autologous stem cell transplantation in multiple myeloma. Blood2012; 119: 687–691.

14 Zhan F, Barlogie B, Arzoumanian V, Huang Y, Williams DR, Hollmig K et al.Gene-expression signature of benign monoclonal gammopathy evident inmultiple myeloma is linked to good prognosis. Blood 2007; 109: 1692–1700.

15 Lahuerta JJ, Mateos MV, Martinez-Lopez J, Rosinol L, Sureda A, de la Rubia J et al.Influence of pre- and post-transplantation responses on outcome of patients withmultiple myeloma: sequential improvement of response and achievement ofcomplete response are associated with longer survival. J Clin Oncol 2008; 26:5775–5782.

16 Rosinol L, Oriol A, Teruel AI, Hernandez D, Lopez-Jimenez J, de la Rubia J et al.Superiority of bortezomib, thalidomide, and dexamethasone (VTD) as inductionpretransplantation therapy in multiple myeloma: a randomized phase 3PETHEMA/GEM study. Blood 2012; 120: 1589–1596.

17 Durie BG, Harousseau JL, Miguel JS, Blade J, Barlogie B, Anderson K et al. Inter-national uniform response criteria for multiple myeloma. Leukemia 2006; 20:1467–1473.

18 Paiva B, Almeida J, Perez-Andres M, Mateo G, Lopez A, Rasillo A et al. Utility offlow cytometry immunophenotyping in multiple myeloma and other clonalplasma cell-related disorders. Cytometry B Clin Cytom 2010; 78: 239–252.

19 Kalina T, Flores-Montero J, van der Velden VH, Martin-Ayuso M, Bottcher S, RitgenM et al. EuroFlow standardization of flow cytometer instrument settings andimmunophenotyping protocols. Leukemia 2012; 26: 1986–2010.

20 Costa ES, Pedreira CE, Barrena S, Lecrevisse Q, Flores J, Quijano S et al. Automatedpattern-guided principal component analysis vs expert-based immunopheno-typic classification of B-cell chronic lymphoproliferative disorders: a stepforward in the standardization of clinical immunophenotyping. Leukemia 2010;24: 1927–1933.

21 Morgan GJ, Walker BA, Davies FE. The genetic architecture of multiple myeloma.Nat Rev Cancer 2012; 12: 335–348.

22 Kuehl WM, Bergsagel PL. Molecular pathogenesis of multiple myeloma and itspremalignant precursor. J Clin Invest 2012; 122: 3456–3463.

23 Weston-Bell N, Gibson J, John M, Ennis S, Pfeifer S, Cezard T et al. Exomesequencing in tracking clonal evolution in multiple myeloma following therapy.Leukemia 2012; 27: 1188–1191.

24 Egan JB, Shi CX, Tembe W, Christoforides A, Kurdoglu A, Sinari S et al.Whole-genome sequencing of multiple myeloma from diagnosis to plasma cellleukemia reveals genomic initiating events, evolution, and clonal tides. Blood2012; 120: 1060–1066.

25 Keats JJ, Chesi M, Egan JB, Garbitt VM, Palmer SE, Braggio E et al. Clonal competitionwith alternating dominance in multiple myeloma. Blood 2012; 120: 1067–1076.

26 Dimopoulos M, Kyle R, Fermand JP, Rajkumar SV, San Miguel J,Chanan-Khan A et al. Consensus recommendations for standard investigativeworkup: report of the International Myeloma Workshop Consensus Panel 3.Blood 2011; 117: 4701–4705.

27 Bergsagel PL, Mateos MV, Gutierrez NC, Rajkumar SV, San Miguel JF. Improvingoverall survival and overcoming adverse prognosis in the treatment ofcytogenetically high-risk multiple myeloma. Blood 2013; 121: 884–892.

28 Avet-Loiseau H, Durie BG, Cavo M, Attal M, Gutierrez N, Haessler J et al. Combiningfluorescent in situ hybridization data with ISS staging improves risk assessment inmyeloma: an International Myeloma Working Group collaborative project.Leukemia 2013; 27: 711–717.

29 Barlogie B, Attal M, Crowley J, van Rhee F, Szymonifka J, Moreau P et al. Long-termfollow-up of autotransplantation trials for multiple myeloma: update of protocolsconducted by the intergroupe francophone du myelome, southwest oncologygroup, and university of arkansas for medical sciences. J Clin Oncol 2010; 28:1209–1214.

30 Mehta J, Singhal S. High-dose chemotherapy and autologous hematopoietic stemcell transplantation in myeloma patients under the age of 65 years. Bone MarrowTransplant 2007; 40: 1101–1114.

31 Martinez-Lopez J, Blade J, Mateos MV, Grande C, Alegre A, Garcia-Larana J et al.Long-term prognostic significance of response in multiple myeloma after stemcell transplantation. Blood 2011; 118: 529–534.

32 Kyle RA, Therneau TM, Rajkumar SV, Offord JR, Larson DR, Plevak MF et al.A long-term study of prognosis in monoclonal gammopathy of undeterminedsignificance. New Engl J Med 2002; 346: 564–569.

33 Kyle RA, Remstein ED, Therneau TM, Dispenzieri A, Kurtin PJ, Hodnefield JM et al.Clinical course and prognosis of smoldering (asymptomatic) multiple myeloma.New Engl J Med 2007; 356: 2582–2590.

34 Cesana C, Klersy C, Barbarano L, Nosari AM, Crugnola M, Pungolino E et al.Prognostic factors for malignant transformation in monoclonal gammopathy ofundetermined significance and smoldering multiple myeloma. J Clin Oncol 2002;20: 1625–1634.

35 Hillengass J, Fechtner K, Weber MA, Bauerle T, Ayyaz S, Heiss C et al. Prognosticsignificance of focal lesions in whole-body magnetic resonance imaging inpatients with asymptomatic multiple myeloma. J Clin Oncol 2010; 28: 1606–1610.

36 Kastritis E, Terpos E, Moulopoulos L, Spyropoulou-Vlachou M, Kanellias N,Eleftherakis-Papaiakovou E et al. Extensive bone marrow infiltration and abnormalfree light chain ratio identifies patients with asymptomatic myeloma at high riskfor progression to symptomatic disease. Leukemia 2013; 27: 947–953.

37 Larsen JT, Kumar SK, Dispenzieri A, Kyle RA, Katzmann JA, Rajkumar SV.Serum free light chain ratio as a biomarker for high-risk smoldering multiplemyeloma. Leukemia 2013; 27: 941–946.

38 Bianchi G, Kyle RA, Larson DR, Witzig TE, Kumar S, Dispenzieri A et al. High levelsof peripheral blood circulating plasma cells as a specific risk factor for progressionof smoldering multiple myeloma. Leukemia 2013; 27: 680–685.

39 Dispenzieri A, Kyle RA, Katzmann JA, Therneau TM, Larson D, Benson J et al.Immunoglobulin free light chain ratio is an independent risk factor for progres-sion of smoldering (asymptomatic) multiple myeloma. Blood 2008; 111: 785–789.

40 Rajkumar SV, Kyle RA, Therneau TM, Melton 3rd LJ, Bradwell AR, Clark RJ et al. Serumfree light chain ratio is an independent risk factor for progression in monoclonalgammopathy of undetermined significance. Blood 2005; 106: 812–817.

41 Katzmann JA, Clark R, Kyle RA, Larson DR, Therneau TM, Melton 3rd LJ et al.Suppression of uninvolved immunoglobulins defined by heavy/light chain pairsuppression is a risk factor for progression of MGUS. Leukemia 2013; 27: 208–212.

42 Perez-Persona E, Vidriales MB, Mateo G, Garcia-Sanz R, Mateos MV, de Coca AG etal. New criteria to identify risk of progression in monoclonal gammopathy ofuncertain significance and smoldering multiple myeloma based on multi-parameter flow cytometry analysis of bone marrow plasma cells. Blood 2007; 110:2586–2592.

43 Paiva B, Gutierrez NC, Chen X, Vidriales MB, Montalban MA, Rosinol L et al. Clinicalsignificance of CD81 expression by clonal plasma cells in high-risk smolderingand symptomatic multiple myeloma patients. Leukemia 2012; 26: 1862–1869.

44 Kumar S, Rajkumar SV. Will the real myeloma please stand up? Leukemia 2013; 27:760–761.

45 Rajkumar SV, Larson D, Kyle RA. Diagnosis of smoldering multiple myeloma. NewEngl J Med 2011; 365: 474–475.

Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)

Flow cytometric MGUS-like profile in myelomaB Paiva et al

2061

& 2013 Macmillan Publishers Limited Leukemia (2013) 2056 – 2061