www.elsevier.com/locate/humpath

Human Pathology (2010) 41, 1702–1710

Original contribution

Immunohistochemical analysis of NOTCH1 and JAGGED1expression in multiple myeloma and monoclonalgammopathy of undetermined significanceAnita Škrtić MDa,⁎, Petra Korać PhDa, Delfa Radić Krišto MDb,Radmila Ajduković Stojisavljević MDc, Davor Ivanković MD, PhDd,Mara Dominis MD, PhDa

aDepartment of Pathology and Cytology, University of Zagreb, School of Medicine, University Hospital Merkur,10000 Zagreb, CroatiabDepartment of Medicine, University Hospital Merkur, 10000 Zagreb, CroatiacDepartment of Medicine, University Hospital Dubrava, 10000 Zagreb, CroatiadDepartment of Medical Statistics, Epidemiology and Medical Informatics, Andrija Stampar School of Public Health,University of Zagreb, School of Medicine, 10000 Zagreb, Croatia

Received 7 December 2009; revised 27 April 2010; accepted 7 May 2010

0d

Keywords:Multiple myeloma;NOTCH1;JAGGED1;IGH translocations;Immunohistochemistry

Summary Notch signaling is implicated in the pathogenesis of multiple myeloma expressing high level ofactive Notch proteins NOTCH1 and JAGGED1 in tumor plasma cells. We investigated expression ofNOTCH1 and JAGGED1 in bone marrow trephine biopsies of 80 newly diagnosed multiple myeloma and20 monoclonal gammopathy of undetermined significance patients using immunohistochemical methods.The number of positive tumor cells was counted per 1000 tumor cells and the intensity of staining wasassessed semi quantitatively. Multiple myelomas expressed NOTCH1 in 92.31% (72/78) and JAGGED1in 92.21% (71/77) cases. NOTCH1 staining was strong in the majority of cases (59.7%), whereasJAGGED1 was predominately weak (67.6% of cases). In contrast, both markers were negative in allmonoclonal gammopathy of undetermined significance cases. However, upon progression of disease frommonoclonal gammopathy of undetermined significance to multiple myeloma (seen in 4 patients), analysisof the subsequent bonemarrow biopsy showed weak expression of bothmarkers in tumorous plasma cells.Immunohistochemistry results were compared with the pattern of bone marrow infiltration, plasma celldifferentiation, and the presence of t(11;14)(q13,q32), t(14;16)(q32;q23),and t(4;14)(p16.3;q23) andoverall survival in multiple myeloma patients. A significant correlation was found between strongNOTCH1 staining in multiple myeloma plasma cells and the diffuse type of bone marrow infiltration (P =.002) and an immature morphologic type of plasma cells (P = .043). After a median follow-up of 20.3months, in multiple myeloma patients no difference in overall survival between NOTCH1 (P = .484) andJAGGED1 (P = .822) positive and negative cases were found. In conclusion, our results indicateimportance of NOTCH1 and JAGGED1 expression in plasma cell neoplasia and a possible diagnosticvalue of their immunohistochemical evaluation of bone marrow infiltrates for multiple myeloma.© 2010 Elsevier Inc. All rights reserved.

⁎ Corresponding author.E-mail address: skrtic.anita@gmail.com (A. Škrtić).

046-8177/$ – see front matter © 2010 Elsevier Inc. All rights reserved.oi:10.1016/j.humpath.2010.05.002

1703NOTCH1 and JAGGED1 expression in multiple myeloma

1. Introduction The involvement of Notch signaling in the development

Plasma cell neoplasms represent variable clinical andbiological entities including more benign forms, such asmonoclonal gammopathy of undetermined significance(MGUS), and more aggressive forms, such as multiplemyeloma (MM) [1]. Multiple myeloma in most patients is, atpresent, an incurable malignancy that is sometimes precededby MGUS.

MGUS and MM are easily clinically distinguishedaccording to diagnostic criteria [2-4] but the biologydifferences and interrelationship between these two entitiesare more complex.

Tumor plasma cells (PC) in MM and MGUS with a self-renewing potential are arrested as intermediates in plasma-cell differentiation. Tumor PC home and distribute to thebonemarrow (BM), where the microenvironment sustains thegrowth and survival of tumor PCs. Usually, in MM, tumorcells are distributed as interstitial clusters, nodules, or diffuseinfiltrates. Tumor PC differentiation varies from mature toimmature cytological forms [5-8]. In MGUS, the well-differentiated, mature morphologic type of PC are distributedas rare interstitial single cells or small clusters within the BM.

Tumor PC in MM and MGUS may carry multiple andcomplex chromosomal abnormalities [9-14]. Currently, 7recurrent IGH translocation partners have been identified inMM and MGUS: CCND1(11q13); CCND2(12p13);CCND3(6p21); MAF(16q23); MAFB(21q12); MAFA(8q24.3), and FGFR3(MMSET)(4p16.3). Most transloca-tions are errors in one of the 3 mechanisms involving B-celldevelopment in the germinal centre: IGH switch recombi-nation, somatic hypermutation and Variable (V) Diversity(D) and Joining (J) genes recombination. These errors mightrepresent one of the first events in the pathogenesis of MM[9,10,12]. These IGH translocations can be detected innearly 50% of MGUS and 55% to 70% of MM [9-14].

Deregulated Notch signaling is also involved in thepathogenesis of MM [15]. NOTCH1 is a member of themammalian Notch family of type I transmembrane receptorsand transcriptional regulators, which play important role inhematopoiesis. Notch signaling is a critical factor involved inthe self-renewal of hematopoietic stem cells, the regulationof T-cell commitment from hematopoietic stem cells, theregulation of αβ T-cell development, and B-cell development[16,17]. Binding of the ligand on an adjacent receptorinduces a conformational change of NOTCH1. Similar toNOTCH1, JAGGED1 is a transmembrane type I protein anda member of mammalian Notch ligand DSL (Delta, Serrate,Jagged) [16,17]. In the absence of JAGGED1 binding, theNOTCH1 receptor is inactive [16,17].

An aberrant truncated constitutively-active NOTCH1receptor, the result of translocation t(7;9)(q34;q34.3), wasfound to be a causative factor in the development of a subsetof T-cell acute lymphoblastic leukaemia/lymphoma [18].Since then, a role for Notch signaling has been found inmany solid and hematopoietic neoplasms [15].

of B-cell neoplasms is less well understood. Both oncogenicand tumor suppressive roles of Notch activity have beenimplicated [15]. This controversial involvement of theNotch signaling pathway in MM comes as a result ofseveral studies of primary and cultured MM cell lines. Somestudies showed Notch-induced growth inhibition andapoptosis. In contrast, some studies showed a proliferationof tumor PC [17,19-24].

Expression of NOTCH1 and JAGGED1 has been seen intumor PC in patients with MM [20]. Both were highlyexpressed in neoplastic CD138 myeloma cells, in contrast tolow expression or no immunohistochemical staining inisolated plasma CD38+++/CD19++ cells from the bonemarrow of healthy donors [20].

We have comparatively analyzed semi quantitativeNOTCH1 and JAGGED1 expression in the tumor PC ofMM and MGUS patients and particularly proteins expres-sion in MM patients who evolved from MGUS. Resultswere compared with PC differentiation, the pattern of BMinfiltration, measures of tumor burden, and the mostfrequent recurrent chromosomal abnormalities characteristicof tumor PC.

To elucidate possible involvement of NOTCH1 andJAGGED1 proteins in the pathogenesis of MGUS, in theprogression of disease from MGUS to MM, in thepathogenesis of MM and their relevance for prognosisprediction, we investigate immunohistochemical analysis ofNOTCH1 and JAGGED1 expression in MM and MGUStumor PC.

2. Patients and methods

2.1. Patients

Eighty consecutive newly diagnosed MM and 20 MGUSpatients from three institutions (University Hospital Merkur;University Hospital Dubrava; University Hospital Osijek)(2004-2009) were included in this study. All patientsfulfilled the diagnostic criteria of plasma cell myeloma orMGUS [2-4]. The median age in MM patients was 68 years(range 35-86); 38 patients were women, and 42 were men. InMGUS patients, the median age was 67 years (range, 43-85);9 patients were women, and 11 were men.

MM patients were analyzed according to measures oftumor burden: age, hemoglobin, creatinine, lactate dehydro-genase, C-reactive protein, serum albumin, serum calcium,bone lesions, serum beta-2 microglobulin, Durie Salmonstage, and immunoglobulin isotype and subtype. The cutoffpoint for clinical and laboratory data was assessed accordingto the study of Greipp et al [6].

According to the Durie Salmon staging system, there were24 MM patients at stage I (30.0%), 18 patients at stage II(22.5%), and 38 patients at stage III (47.5%) [25].

1704 A. Škrtić et al.

MM patients were treated in accordance with theguidelines on the management of multiple myeloma fromthe UK Myeloma Forum, Nordic Myeloma Study Groupand British Committee for Standards in Haematologypublished in 2001 and 2005 [26,27] (http://www.myeloma-nordic.org). Forty-seven MM patients were treated accord-ing to the guidelines on the management of non-transplanteligible patients, whereas 21 MM patients were treatedaccording to the guidelines on the management of eligiblepatients for autologous stem cell transplantation. TwelveMM patients and all MGUS patients were monitored withregular clinical assessment and laboratory follow-up.

The median time during which disease in four MGUSpatients progressed and fulfilled the diagnostic criteria forMM was 17.65 months (range, 3.7-29.5 months). Two ofthese MM patients needed only regular monitoring, and theother 2 were treated according to guidelines on themanagement of non–transplant-eligible patients.

The median follow-up was 18.8 months in MM patientsand 20.7 months in MGUS patients.

The study was approved by the University HospitalMerkur ethical committee.

2.2. Methods

Bone marrow trephine biopsies were performed at thetime of the diagnosis. Formalin fixed, decalcified (Osteosoft,Merk), paraffin-embedded (FFPE) bone marrow biopsieswere processed routinely and evaluated at UniversityHospital Merkur. Giemsa stain was preformed to betterevaluate the morphology of the PC.

Sequentially sectioned 4-μm-thick slides were used forperforming immunohistochemical staining in each analyzedcase. Immunohistochemical staining was performed by anautomated immunostainer (Dako Autostainer Plus, Dako-Cytomation, Glostrup, Denmark) using LSAB HRP andHRP+ kits according to the manufacturer's instructions. Forestablishing the diagnosis of MGUS and MM, tumor PCwere immunohistochemically stained for CD138 (clone5F7, Novocastra, Leica Biosystems Newcastle Ltd., UK,dilution 1:25) and CD38 (clone 38C03, LabVison, ThermoFisher Scientific, Fremont, CA, USA dilution 1:50),according to the manufacture's instructions, to identify PCby membranous and cytoplasmic staining.

The goat polyclonal anti-NOTCH1 antibody (clone sc-6014,Santa Cruz Biotechnology) was used at 1:400 dilution. Therabbit polyclonal anti-JAGGED1 antibody (clone sc-6011;Santa Cruz Biotechnology) was used at 1:150 dilution.A positive control for NOTCH1 and JAGGED1 expression(cytoplasmic staining) was samples of lung adenocarcinoma.For the negative control, adjacent sections were stained in theabsence of a primary antibody. Subsequently, immunostainedslides were analyzed for each antibody.

To accurately enumerate few PC in MGUS, in eachMGUS and MM case, we compared the results of CD138and CD38 immunohistochemical staining in tumor PC with

the counted number of morphologically observed tumor PCin immunohistochemically stained NOTCH1 and JAGGED1sequential slides. We found no difference in observing thepercentage of PC in bone marrow samples of both MGUSand MM cases. Simultaneous double-labeled immunohisto-chemical staining using NOTCH1 or JAGGED1 and CD138was not performed.

For each slide, adjacent digital images of tumor represen-tative fields were acquired, using an Olympus 71 digitalcamera and an Olympus BX51 microscope with an ×40objective and saved as uncompressed 24-bit RGB TIFF files.The field size was 144436.18 μm2 and was measured in thesoftware program AnalySIS (Olympus Soft Imaging SystemGmbH, Munster, Germany). Representative digital imageswere analyzed in each slide to count 1000 tumor PC using the“Grid” software function, dividing images into 10 subfieldsmeasuring 50 × 50 μm. Over 1000 positively immunolabeledmalignant PC were counted. The ratio between positivelylabeled tumor cells and the 1000 counted tumor PC wascalculated. The median value of the calculated ratio ± 1 SDwas considered as the cutoff point for determining positivity(NOTCH1, median ± 1 SD, 28%; JAGGED1, median ± 1 SD,25%). Cases were considered positive for NOTCH1 andJAGGED1 expression if at least 30% of the tumor cells hadcytoplasmatic positivity. A cutoff point of 30% was useful todetermine MM from MGUS cases, avoiding false-positiveresults and immunolabelling of phenotypically normalpolyclonal PCs in MGUS, which account for an average of27% according to the study of Olteanu et al [28].

The intensity of immunostaining was semi quantified andlabeled as “strong,” “weak,” or “no staining.” “Weak”staining was considered when the intensity of staining wasvisible only with a ×40 objective [29].

Fluorescent in situ hybridization was performed on 4-μm-thick, FFPE BM trephine sections according to the pre-viously described protocol [30] (http://jmd.amjpathol.org).

The IGH/CCND1, IGH/MAF, and IGH/FGFR3(MM-SET) translocations were analyzed with dual-color fusionprobes (Vysis).

FFPE BM trephines were routinely and immunohisto-chemically stained and then evaluated by two independentobservers (A.Š. and M.D.), and FISH results were interpretedby three independent observers in a blind study (A.Š., M.D.and P.K.). Interobserver subjective score reproducibility inFISH signal interpretation was good (91%). In nine MMcases, differences in scoring of the FISH results between thethree observers occurred, and the score was re-evaluated toreach consensus between the observers.

2.3. Statistical analysis

The χ2 test was used to compare the differences inNOTCH1 or JAGGED1 expression and PC morphologiccharacteristics and chromosomal abnormalities. Kaplan–Meier survival analysis was performed to estimate the overallsurvival (OS) time within MM and MGUS as well as OS

Table 1 Morphologic type of plasma cell differentiation andtype of bone marrow infiltration in multiple myeloma,monoclonal gammopathy of undetermined significance, andmultiple myeloma evolved from monoclonal gammopathy ofundetermined significance patients

PC differentiation BM infiltration

Bartl'sgrade I/II

Bartl'sgrade III

Diffuse Nondiffuse

MM 66 14 31 49MGUS 20 0 0 20MM evolved fromMGUS

4 0 0 4

1705NOTCH1 and JAGGED1 expression in multiple myeloma

time within multiple myeloma NOTCH1 and JAGGED1positive and negative subgroups. The OS times of subgroupswere compared using the log-rank test. Overall survival timerepresents the time from the date of diagnosis until the lastfollow-up visit or death. The differences were consideredsignificant if the p value was less than 5% (P b .05).Statistical analyses were performed with STATISTICA dataanalysis software system [version 8.0, www.statsoft.com,StatSoft, Inc (2008), Tulsa, OK, USA].

Fig. 1 Multiple myeloma bone marrow morphology. Bone marrow semarrow aspirate smears with (C) Bartl's grade III of PC (D) Bartl's grad

3. Results

3.1. Morphology

A total of 31/80 MM patients had diffuse BM infiltrationand 49 of 80 cases had a non-diffuse type of BM infiltration.An immature morphologic type of tumor PCs (Bartl's gradeIII) was found in 14 of 80 MM cases, whereas 66/80 of MMcases had mature morphologic forms of tumor PC (Bartl'sgrade I/II) [6,8] (Table 1, Fig. 1). All MGUS cases showed anon-diffuse type of BM infiltration with a mature morpho-logic type (Bartl's grade I) of PC (Table 1). Four MM caseswith previously documented MGUS showed a nondiffusetype of BM infiltration with a mature morphologic type ofPC (Bartl's grade I) (Table 1).

3.2. Immunohistochemistry

In the MM group, immunohistochemical analysis showedNOTCH1 expression in 72 (92.31%) of 78 (range ofNOTCH1 expression in PCs = 30%-94%, mean 79%) andJAGGED1 in 71 (92.21%) of 77 PC (range of JAGGED1expression in PCs = 30%-84%, mean 53%) (Table 2, Fig. 2).

ctions with pattern of infiltration (A) diffuse (B) nondiffuse. Bonee I of myeloma cell.

Table 2 Immunohistochemistry and analyzed cytogenetic translocations inmultiple myeloma, monoclonal gammopathy of undeterminedsignificance, and multiple myeloma evolved from monoclonal gammopathy of undetermined significance patients

MM MGUS MM evolved from MGUS

n = 80 n = 20 n = 4

+ − NA + − NA + − NA

NOTCH1 expression 72 6 2 0 20 0 4 0 0JAGGED1 expression 71 6 3 0 20 0 4 0 0IGH/CCND1 11 68 1 3 17 0 0 4 0IGH/FGFR3(MMSET) 20 59 1 4 16 0 1 3 0IGH/MAF 5 74 1 1 19 0 1 3 0

FISH, fluorescence in situ hybridization; NA, not analyzed (poor sample quality); +, positive immunohistochemical staining or present cytogenetictranslocation; −, negative immunohistochemical staining or no cytogenetic translocation.

1706 A. Škrtić et al.

NOTCH1 expression in 6 (7.7%) of 78 MM cases, as wellas JAGGED1 expression in 6 (7.8%) of 77 MM cases wereconsidered negative according to the cut off level of 30% forimmunohistochemical staining (range of NOTCH1 expres-sion in PC = 0.3%-28%, mean 18%; range of JAGGED1expression in PC = 2.7%-26%, mean 20%). Results forimmunohistochemical analysis of NOTCH1 could not beinterpreted in two MM and JAGGED1 expression in 3 MM

Fig. 2 Immunohistochemical staining of NOTCH1 and JAGGED1 in mstrong (A and C); weak, (B and D).

cases due to the poor quality of performed and repeatedimmunohistochemical staining.

In the MGUS group, we found NOTCH1 and JAGGED1protein expression in tumor cells, but it was below thecutoff value to consider it positive (range of NOTCH1expression in PC = 0.5%-9.6%, mean 3.2%; range ofJAGGED1 expression in PC = 0.8%-10.4%, mean 2.3%)(Table 2).

ultiple myeloma. Intensity of NOTCH1 and JAGGED1 expression;

Table 4 Semiquantified immunohistochemical staining ofNOTCH1 and JAGGED1 expression, diffuse type of bonemarrow infiltration, and Bartl's grade III of PC differentiation inmultiple myeloma

Staining intensity BM infiltrationdiffuse type

PCdifferentiationBartl's grade III

n χ2 n χ2

NOTCH1 Strong 22 5.263 11 4.087Weak 7 (P = .0022) 2 (P = .0431)

JAGGED1 Strong 13 2.830 5 0.263Weak 17 (P = .0920) 8 (P = .6050)

1707NOTCH1 and JAGGED1 expression in multiple myeloma

Analyzing a first biopsy of 4 (20%) of 20 MGUS casesthat progressed to MM, tumor PC showed NOTCH1 andJAGGED1 positivity but were also bellow the value toconsider them positive (range of NOTCH1 expression inPC = 3%-8% mean 4.7%; range of JAGGED1 expressionin PC = 1.5%-7.7%, mean 4.1%). In contrast, in a secondBM biopsy of these four cases when they fulfilled thediagnostic MM criteria positivity of NOTCH1 andJAGGED1 labeled tumor PCs according to cutoff pointwas found (range of NOTCH1 expression in PC = 59%-86.8% mean 76%; range of JAGGED1 expression in PC =67.1%-86.9%, mean 78%) (Table 2).

Immunohistochemical semi quantified analysis ofNOTCH1 expression in MM tumor cells showed strongstaining in 43 (59.7%) of 72 and weak staining in 29 (40.3%)of 72 cases. Immunohistochemical analysis of JAGGED1expression showed strong staining in 23 (32.4%) of 71 andweak staining in 48 (67.6%) of 71 MM cases (Table 3).

Tumor PC showed weak staining of both NOTCH1 andJAGGED1 proteins in all four MM cases that evolved fromMGUS (Table 3).

Weak NOTCH-1 and JAGGED-1 immunostaining wasfound in megakaryocytes, and in some BM microenvi-ronmental network cells such as adventitial reticular cells,perisinusoidal and periarterial adventitial cells, andcapillary endothelial cells. BM microenvironmental net-work cells showed an irregular immunostaining patternwith some negative cells present. Since megakaryocyteswere invariably NOTCH1 and JAGGED1 immunolabeled,they could be used as an internal control for the qualityof staining.

3.3. FISH

3.3.1. Chromosomal rearrangements of IGH/CCND1,IGH/MAF, and IGH/FGFR3(MMSET)

InMM, translocation IGH/FGFR3(MMSET)was observedin 20 (25.3%) of 79 cases; IGH/CCND1, in 11 (13.9%) of 79cases; and IGH/MAF, in 5 (6.4%) of 78 cases (Table 2).

In MGUS, IGH/FGFR3(MMSET) was found in 4 (20%)of 20 cases; IGH/CCND1, in 3 (15%) of 20 cases; and IGH/MAF, in 1 (5%) of 20 cases (Table 2).

Table 3 Semiquantified immunohistochemical staining ofNOTCH1 and JAGGED1 expression in tumor PC of multiplemyeloma and multiple myeloma evolved from monoclonalgammopathy of undetermined significance

Strong Weak NA

MM (n = 80)NOTCH1 expression 43 29 2JAGGED1 expression 23 48 3MM evolved from MGUS (n = 4)NOTCH1 expression 0 4 0JAGGED1 expression 0 4 0

In MM cases that evolved from MGUS, we found inbiopsies, both, the first and the second, translocation IGH/FGFR3(MMSET) in 1 (25%) of 4 cases and IGH/CCND1 in1 (25%) of 4 cases with no additional changes in the analyzedchromosomal rearrangements (Table 2).

A statistically significant difference was found betweenthe strong intensity of NOTCH1 staining in MM PC, thediffuse pattern of BM infiltration (P = .0022) and theimmature cytologic type of tumor PC (P = .0431) (Table 4).

Comparison of the NOTCH1 and JAGGED1 immuno-histochemical staining intensity showed no statisticalsignificance (P = .184) in MM patients.

The NOTCH1 and JAGGED1 expression in MM PCcompared to analyze chromosomal aberrations also showedno statistical significance.

Overall survival times of MM and MGUS patients werecompared. The MGUS group of patients showed astatistically significant longer overall survival (median 33.8months for MGUS patients versus median 20.7 months forMM patients (P = .014) (Fig. 3). The presence of NOTCH1or JAGGED1 expression was not associated with MM

Fig. 3 Overall survival distribution (Kaplan-Meier) of MM andMGUS patients Circles represents complete follow-up; plusesrepresent censored data (log-rank test = 2.8241 P = .014); full linerepresents MM patients; broken line represents MGUS patients.

Fig. 5 Overall survival distribution (Kaplan-Meier) of patientswith positive or negative JAGGED1 expression in myeloma cellsCircles represents complete follow-up; pluses represents censored

1708 A. Škrtić et al.

overall survival (NOTCH1 P = .484; JAGGED1 P = .822)(Figs. 4, 5).

NOTCH1 and JAGGED1 expression in MM caseswas also not associated with measures of tumor burden:age (NOTCH1 P = .162; JAGGED1 P = .246); hemoglobin(NOTCH1, P = .635; JAGGED1, P = .441); creatinine(NOTCH1, P = .374; JAGGED1, P = .644); LDH (NOTCH1,P = .748; JAGGED1 P = .118); CRP (NOTCH1, P = .584;JAGGED1, P = .564); serum albumin (NOTCH1, P = .473;JAGGED1, P = .528); serum calcium (NOTCH1, P =.452; JAGGED1, P = .182); bone lesions (NOTCH1, P =.200; JAGGED1, P = .559); serum beta2 microglobulin(NOTCH1, P = .184; JAGGED1, P = .617); Durie Salmonstage (NOTCH1, P = .151; JAGGED1, P = .285);immunoglobulin isotype (NOTCH1, P = .567; JAGGED1,P = .527) and immunoglobulin subtype (NOTCH1, P =.162; JAGGED1, P = .405).

data. (log-rank test = 0.3588; P = .822); full line representsJAGGED1 negative myeloma cells; broken line representsJAGGED1 positive myeloma cells.

4. Discussion

Our results showed the occurrence of NOTCH1 andJAGGED1 expression in MM cells, but we have found nosignificant expression of those proteins in MGUS cells. Inaddition, we showed different occurrence of NOTCH1 andJAGGED1 proteins in PC in MM evolving from MGUS,which showed an advantage of NOTCH1 and JAGGED1expression in MM PC over their expression in MGUS PC.

Regarding the phenotypic profile of tumor PC in MGUSand MM Olteanu et al [28] performed immunophenotypicanalysis using four-color flow cytometry on 32 MM and 32MGUS BM specimens demonstrating in all MGUS casestwo distinct subpopulations of PC, one phenotypicallynormal and polyclonal and the other phenotypically aberrant

Fig. 4 Overall survival distribution (Kaplan-Meier) of patientswith positive or negative NOTCH1 expression in myeloma cells.Circles represent complete follow-up; pluses represents censoreddata (log-rank test = 0.7511; P = .484); full line representsNOTCH1 positive myeloma cells; broken line represents NOTCH1negative myeloma cells.

and monoclonal. Normal PC accounted for an average of27% in total PC count [28].

In our study, the MGUS group showed NOTCH1 andJAGGED1 expression mean values less than 4% of countedPCs per case, which is because of the cut-off values forimmunohistochemical staining were too low for consideringpositive. The number of weak NOTCH1 and JAGGED1positive PCs in MGUS was too small to be explained by thephenotypic heterogeneity of PCs. In addition, NOTCH1 andJAGGED1 expression could not confirm the presence ofdifferent PC subpopulations in both MGUS and MM.According to these data, the cutoff point of 30% was usefulto determine MM from MGUS cases, avoiding false-positiveresults and immunolabelling of phenotypically normalpolyclonal PC in MGUS according to the study of Olteanuet al [28].

Simultaneous double-labeled immunohistochemicalstaining using NOTCH1 or JAGGED1 and CD138 was notperformed. Sequentially sectioned immunohistochemicallystained slides for each investigated protein presented areliable method for analyzing NOTCH1 and JAGGED1protein expression in tumor PC in both MGUS and MM.This is particularly important for MGUS cases due to the lowpercentage of PC (b10%). However, the morphology ofthese tumor cells, because they are well-differentiated cells,presents additional help in validating of NOTCH1 andJAGGED1 immunohistochemical staining in MGUS cases.

We confirmed the almost invariable occurrence ofNOTCH1 and JAGGED1 expression in MM PC withvarious intensity and invariable occurrence in MM evolvedfrom MGUS. Compared to a previously published study[20], our results showed a slight discrepancy. Until now,only Jundt et al [20] had reported a high expression ofNOTCH1 and JAGGED1 in CD138 positive tumor PCusing immunohistochemistry in FFPE BM trephine biopsies

1709NOTCH1 and JAGGED1 expression in multiple myeloma

of 16 newly diagnosed MM patients. In contrast, isolatednormal CD38+++/CD19++ PCs from the BM of healthydonors showed low or no expression of NOTCH1 andJAGGED1 [20].

We showed invariable NOTCH1 and JAGGED1expression in MM patients with a documented history ofMGUS by analyzing a second BM biopsy. This was incontrast to the results of first biopsy, where tumor PCshowed NOTCH1 and JAGGED1 expression at a valuethat was below the value to consider positive. This canindicate NOTCH1 and JAGGED1 involvement in plasmacell tumorigenesis.

In our semiquantitative immunohistochemical analysis ofNOTCH1 and JAGGED1 expression in tumor PCs we founda minority of MM cases (around 8%) with no NOTCH1 andJAGGED1 expression according to the cutoff level. Three ofthese negative MM cases for both investigated proteins haddistinct types of BM infiltration, but all 3 showed maturemorphologic type of tumor PC and the occurrence of IGH/FGFR3(MMSET) translocation in 2 of 3 cases. All 3 MMpatients were males, older than 65 years, at Durie-Salmonstage I and III, and treated with therapy for non–transplant-eligible patients. The median follow-up for these patientswas 27.26 months (range, 0.23-34.5 months).

Various intensities of NOTCH1 and JAGGED1 immu-nohistochemical staining in tumor PCs of most MM cases,not described earlier, was considerable, thus suggestingdifferent levels of protein expression between the cases.Statistical significance was found between the strongintensity of NOTCH1 staining in MM cells and the highgrade of PC differentiation (P = .0022) and diffuse type ofBM infiltration (P = .0431). To estimate the algorithm and topresent clinical applications of the survival data for MMpatients Greipp et al [6] used an extreme regression statisticalmodel, which we did not use due to the small sample of data.Numerous studies about multiple myeloma identified anumber of clinical and laboratory parameters that wereindependent predictors of survival duration [6]. We found nostatistical significance between NOTCH1 and JAGGED1expression in MM cases and measures of tumor burdenincluding age, hemoglobin, creatinine, LDH, CRP, serumalbumin, serum calcium, bone lesions, serum beta2 micro-globulin, Durie Salmon stage, immunoglobulin isotype, andimmunoglobulin subtype.

After a median follow-up of 20.3 months in MM patients,no difference in overall survival between NOTCH1 (P =.484) and JAGGED1 (P = .822) positive and negative caseswere found.

The Notch signaling pathway in malignant B cells hasbeen already described in several studies. Two studies[21,24] showed growth arrest and apoptosis in humanmyeloma-derived cell lines by activating the Notch pathway.In the study of Nefedova et al [21], a unique functionattributed exclusively to NOTCH1 in human myeloma-derived cell lines was described, suggesting the involvementof the Notch signaling pathway in MM resistance to

chemotherapy in which the BM microenvironment plays acritical role [21]. In contrast, 2 studies by Jundt et al found aproliferative effect of the activated Notch signaling pathwayin human myeloma-derived cell lines. They were not able toexclude NOTCH1 as a marker of a particular cell type ofdifferentiation stage in PCs [20].

Taken together, whether the presence of NOTCHsignaling pathway has a proliferative or growth arrest effector whether it plays a role in the complex network of the BMmicroenvironment, it remains clear that both NOTCH1 andJAGGED1 proteins play a role in plasma cell tumorigenesis.The lack or presence of their weak expression in MGUS PCand their expression in MM cases evolved from MGUSindicates their possible involvement in plasma cell tumori-genesis. Therefore, both NOTCH 1 and JAGGED1 expres-sion could possibly suggest more complex changes in PCand, maybe, if confirmed in a larger patient series, serve aspotential markers for MM stages.

References

[1] Grogan TM, Van Camp B, Kyle RA, Muller-Hermelink HK, HarrisNL. Plasma cell neoplasms. In: Jaffe ES, Harris NL, Stein H,Vardiman JW, editors. World Health Organization classification oftumours: pathology and genetics of tumours of haematopoietic andlymphoid tissues. Lyon (FR): IARC Press; 2001.

[2] Durie BG. Cellular and molecular genetic features of myelomaand related disorders. Hematol Oncol Clin North Am 1992;6:463-77.

[3] Salmon SE, Seligmann M. Plasma cell neoplasms. In: De Vita VT,Hellman S, Rosenberg S, editors. Cancer, principles and practice ofOncology. Philadelphia: J.B. Lippincott; 1998. p. 1854.

[4] International Myeloma Working Group. Criteria for the classificationof monoclonal gammopathies, multiple myeloma and related dis-orders: a report of the International Myeloma Working Group. Br JHaematol 2003;121:749-57.

[5] Bartl R, Frisch B, Burkhardt R, et al. Bone marrow histology inmyeloma: its importance in diagnosis, prognosis, classification andstaging. Br J Haematol 1982;51:361-75.

[6] Greipp PR, Leong T, Bennett JM, et al. Plasmablastic morphology–anindependent prognostic factor with clinical and laboratory correlates:Eastern Cooperative Oncology Group (ECOG) myeloma trial E9486report by the ECOG Myeloma Laboratory Group. Blood 1998;91:2501-7.

[7] Bartl R, Frisch B, Fateh-Moghadam A, Kettner G, Jaeger K,Sommerfeld W. Histologic classification and staging of multiplemyeloma. A retrospective and prospective study of 674 cases. Am JClin Pathol 1987;87:342-55.

[8] Brunning RD, McKenna RW. Plasma cell dyscrasias and relateddisorders. Atlas of Tumor Pathology. Washington, D.C.: ArmedForces Institute of Pathology; 1994. p. 323-69.

[9] Chesi M, Bergsagel PL, Brents LA, Smith CM, Gerhard DS, KuehlWM. Dysregulation of cyclin D1 by translocation into an IgH gammaswitch region in two multiple myeloma cell lines. Blood 1996;88:674-81.

[10] González D, van der Burg M, García-Sanz R, et al. Immunoglobulingene rearrangements and the pathogenesis of multiple myeloma. Blood2007;110:3112-21.

[11] Bergsagel PL, Kuehl WM. Critical roles for immunoglobulintranslocations and cyclin D dysregulation in multiple myeloma.Immunol Rev 2003;194:96-104.

1710 A. Škrtić et al.

[12] Bergsagel PL, Kuehl WM, Zhan F, Sawyer J, Barlogie B, ShaughnessyJr J. Cyclin D dysregulation: an early and unifying pathogenic event inmultiple myeloma. Blood 2005;106:296-303.

[13] Avet-Loiseau H, Facon T, Grosbois B, et al. Intergroupe Francophonedu Myélome. Oncogenesis of multiple myeloma: 14q32 and 13qchromosomal abnormalities are not randomly distributed, but correlatewith natural history, immunological features, and clinical presentation.Blood 2002;99:2185-91.

[14] Fonseca R, Bailey RJ, Ahmann GJ, et al. Genomic abnormalities inmonoclonal gammopathy of undetermined significance. Blood 2002;100:1417-24.

[15] Leong KG, Karsan A. Recent insights into the role of Notch signalingin tumorigenesis. Blood 2006;107:2223-33.

[16] Milner LA, Bigas A. Notch as a mediator of cell fate determination inhematopoiesis: evidence and speculation. Blood 1999;93:2431-48.

[17] Artavanis-Tsakonas S, Rand MD, Lake RJ. Notch signaling: cell fatecontrol and signal integration in development. Science 1999;284:770-6.

[18] Ellisen LW, Bird J, West DC, et al. TAN-1, the human homolog of theDrosophila notch gene, is broken by chromosomal translocations in Tlymphoblastic neoplasms. Cell 1991;66:649-61.

[19] Houde C, Li Y, Song L, et al. Overexpression of the NOTCH ligandJAG2 in malignant plasma cells from multiple myeloma patients andcell lines. Blood 2004;104:3697-704.

[20] Jundt F, Pröbsting KS, Anagnostopoulos I, et al. Jagged1-inducedNotch signaling drives proliferation of multiple myeloma cells. Blood2004;103:3511-5.

[21] Nefedova Y, Cheng P, Alsina M, Dalton WS, Gabrilovich DI.Involvement of Notch-1 signaling in bone marrow stroma-mediated denovo drug resistance of myeloma and other malignant lymphoid celllines. Blood 2004;103:3503-10.

[22] Morimura T, Goitsuka R, Zhang Y, Saito I, Reth M, Kitamura D. Cellcycle arrest and apoptosis induced by Notch1 in B cells. J Biol Chem2000;275:36523-31.

[23] Romer S, Saunders U, Jäck HM, Jehn BM. Notch1 enhances B-cellreceptor-induced apoptosis in mature activated B cells withoutaffecting cell cycle progression and surface IgM expression. CellDeath Differ 2003;10:833-44.

[24] Zweidler-McKay PA, He Y, Xu L, et al. Notch signaling is a potentinducer of growth arrest and apoptosis in a wide range of B-cellmalignancies. Blood 2005;106:3898-906.

[25] Durie BG, Salmon SE. A clinical staging system for multiplemyeloma. Correlation of measured myeloma cell mass with presentingclinical features, response to treatment, and survival. Cancer 1975;36:842-54.

[26] Smith A, Wisloff F, Samson D, UK Myeloma Forum, NordicMyeloma Study Group, British Committee for Standards in Haema-tology. Guidelines on the diagnosis and management of multiplemyeloma 2005. Br J Haematol 2006;132:410-51.

[27] UK Myeloma Forum, British Committee for Standards in Haematol-ogy. Diagnosis and management of multiple myeloma. Br J Haematol2001;115:522-40.

[28] Olteanu H, Wang HY, Chen W, McKenna RW, Karandikar NJ.Immunophenotypic studies of monoclonal gammopathy of undeter-mined significance. BMC Clin Pathol 2008;8:13.

[29] Cook JR, His ED, Worley S, Tubbs RR, Hussein M. Immunohisto-chemical analysis identifies two cyclin D1+ subsets of plasma cellmyeloma, each associated with favorable survival. Am J Clin Pathol2006;125:615-24.

[30] Ventura RA, Martin-Subero JI, Jones M, et al. FISH analysis for thedetection of lymphoma-associated chromosomal abnormalities inroutine paraffin-embedded tissue. J Mol Diagn 2006;8:141-51.