Increased vascularization in myeloma

Laroche M, Brousset P, Ludot I, MazieÁres B, Thiechart M, Attal M.Increased vascularization in myeloma.Eur J Haematol 2001: 66: 89±93. # Munksgaard 2001.

Abstract: Several studies have shown that the number of intratumoralvessels can predict the aggressiveness of a solid cancer, development ofmetastases and patient survival. Does angiogenesis play an importantrole in myeloma?The aim of our study was to quantify bone marrow vascularity in variousstages of proliferative plasma cell disorders (monoclonal gammopathiesof undetermined signi®cance (MGUS), stage I multiple myeloma (MM),stage III MM and WaldenstroÈm's macroglobulinemia (WM)) and tocompare it with that of patients with osteoporosis.The study included 15 MGUS patients, 15 patients with stage I MM, 15patients with symptomatic stage III MM, 7 patients with WM, 10patients with osteoporosis, and 10 patients with reactive bone marrow(RBM), matched for sex and age. After iliac crest biopsy, the variousvessels (arterioles, capillaries and sinusoids) were labeled with mono-clonal antibodies CD34 and counted, and a histomorphometric studywas done. The number of arterioles and arterial capillaries wassigni®cantly increased in MGUS and myeloma compared withosteoporosis. The number of arterioles and arterial capillaries increasedmoderately according to the stage of gravity of myeloma. The number ofarterioles is negatively correlated with the trabecular bone volume andpositively correlated with the eroded surfaces.

M. Laroche1, P. Brousset2, I. Ludot3,B. MazieÁ res1, M. Thiechart3,M. Attal4

1Service de Rhumatologie, CHU Rangueil, Toulouse,

France,2Service d'Anatomopathologie, CHU Purpan,

Toulouse, France,3Laboratoire de Recherche en

Pathologie OsteÂo-Articulaire, CHU Rangueil, Toulouse,

France and4Service d'HeÂmatologie, CHU Purpan,

Toulouse, France

Key words: monoclonal gammopathies of

undetermined signi®cance (MGUS); myeloma;

increased vascularity

Correspondence: Dr M. Laroche, Service de

Rhumatologie, CHU Rangueil, 1 avenue Jean-PoulheÁs,

F-31403 Toulouse Cedex 4, France

Tel: +33.5.61.32.27.25

Fax: +33.5.61.32.29.34

Accepted for publication 27 September 2000

In many solid tumors (kidney, prostate, breast) ithas been shown that angiogenesis is an indispensableprerequisite for tumor development and metastasis(1). Several works have shown that aggressiveness,development of metastases and patient survivalcould be predicted by the number of intratumoralvessels (2). Growth factors as vascular endothelialgrowth factor (VEGF), basic ®broblast growthfactor (BFGF), epidermal growth factor receptor(EGFR), tumor angiogenic factor (TAF) andvascular permeability factor (VPF) are thought tobe involved in tumoral angiogenesis and to affect theprognosis of these tumors (3).

Normal or tumoral plasma cells are often found tobe distributed around capillaries in the hemapoieticmarrow. Do malignant plasma cell proliferationshave increased vascularity? Does angiogenesis playan important role in myeloma (MM)? Our studyaimed to quantify the vascularity of various stages ofplasma cell proliferation (monoclonal gammopa-

thies of undetermined signi®cance (MGUS), stage Iand stage III MM) and to compare this vascularitywith that of patients with osteoporosis and patientswith lymphocytic proliferation of WaldenstroÈm'smacroglobulinemia (WM).

Patients and methods

Patients

The study comprised: 15 patients with MGUSaccording to Kyle's de®nition (4): 6 women, 4men, 7 IgG, 3 IgA, mean age 58.6 yr (range 47±74 yr); 15 patients with asymptomatic stage Imyeloma according to the de®nition of Durie andSalmon (5): 6 women, 4 men, 7 IgG, 3 IgA, mean age54.8 yr (range 43±72 yr); 15 patients with sympto-matic stage IIIa myeloma according to Durie andSalmon: 6 women, 4 men, 7 IgG, 3 IgA, mean age61 yr (range 46±70 yr); 10 patients with post-

Eur J Haematol 2001: 66: 89±93Printed in UK. All rights reserved

Copyright # Munksgaard 2001

EUROPEANJOURNAL OF HAEMATOLOGY

ISSN 0902-4441

89

menopausal or `idiopathic' osteoporosis: 6 women, 4men, mean age 59 yr (range 45±70 yr); these bonebiopsies were done before 1980, before the use of x-ray absorptiometry, when histomorphometry wasuseful for diagnosis of osteoporosis; 10 patients withreactive bone marow (RBM): 6 women, 4 men, meanage 61 yr (range 37±75 yr): bone biopsies were donefor an unaccounted in¯ammatory biological syn-drome; 7 patients with WaldenstroÈm's macroglobu-linemia: 5 women, 2 men, mean age 62 yr (range 52±72 yr).

Methods

The MGUS and myeloma patients were matchedfor age, sex and type of monoclonal immunoglo-bulin. All patients were investigated before anytreatment for their disorder. They were matched forage, sex and percentage of resorption surfaces withthe patients with MGUS or myeloma. Reactivebone marrow patients were matched for age, sexand bone marrow cellularity with MGUS or stage 1myeloma patients.

Histologic assessment

The number of marrow plasma cells was determinedin all patients by bone marrow count after sternalaspiration.

Patients underwent transiliac bone biopsy using aBordier's trephine (8 mm internal diameter).Biopsies were ®xed in 5% formalin, dehydratedwith alcohol and embedded in methylmethacrylatewithout decalci®cation. Seven-micrometre sectionswere cut using a Jung microtome and stained withSolochrom cyanin and Goldner trichrome.Quantitative bone histomorphometry was under-taken on cancellous bone using a semi-automaticsystem (Biocom, Kontrom Instruments) and wasexpressed according to the recommendations of theAmerican Society for Bone and Mineral Research(6): bone trabecular volume (BTV), the proportionof tissue volume occupied by both mineralized andunmineralized bone; eroded surfaces (ES), theproportion of the bone surfaces occupied byerosions assumed to be due to the activity ofosteoclasts; osteoid surfaces (OS), the proportionof bone surfaces occupied by unmineralized bone;osteoid volume (OV) the proportion of bone volumeoccupied by unmineralized bone. Bone histomor-phometry was not done for reactive bone marrowpatients.

Blood vessels were highlighted by staining endo-thelial cells for CD34 monoclonal antibodies(QBENd10, Immunotech, Marseille, France) usinga standard peroxidase staining technique asdescribed previously (7). For each patient, 8 ®elds

per slide and 4 slides were examinated withmagni®cation r200. The numbers of vessels(arterioles and capillaries) were mathematicallycalculated in mm2. The vessels were quanti®ed inall hematopoietic marrow and not only in theplasmocytoma nodes (hot spots). Bone marrowcellularity (the percentage of the surface occupied byhematopoietic cells), the percentage of the surfaceoccupied by adipocytes and the percentage of thesurface occupied by sinusoids (venous capillaries)were determined using an integrated eyepiece.

The arterioles, capillaries and sinusoids arecharacterized according to the classi®cation ofBurkhardt (8, 9): the arterioles are characterizedby a wall of 2±3 sheets of muscle cells, an adventiceand a diameter 5 mm. The capillaries are character-ized by a wall composed of a single sheet of musclecells and an endothelium. The sinusoids arecharacterized by a simple endothelium wall withoutany muscle cells.

Histomorphometric parameters were analyzedand vessel counts performed by two investigators,and the mean of the two counts was calculated andexpressed as the result. The two investigatorsworked independently. The inter-observer varia-tions, determined on 10 slides, was 8% for arterioles,10% for capillaries, and 5% for sinusoids.

Statistical analysis

The parameters of the various groups werecompared using an Anova test, and correlationswere sought using Spearman's test.

Results (Tables 1 and 2)

No difference was observed between the variousgroups according to age and sex.

Bone marrow cellularity was signi®cantlyincreased in stage I and stage III myelomas and inWaldenstroÈm's macroglobulinemia compared withosteoporosis.

The number of intramedullary arterioles wassigni®cantly increased in all stages of plasma celldisorder and in WaldenstroÈm's disease comparedwith osteoporosis. Similarly, the number of arterialcapillaries was much greater in MGUS and inmyeloma. The number of arterioles and capillariesincreased moderately according to the stage andgravity of plasma cell disorder. The number of intra-medullary arterioles and capillaries of RBM patientswas situated between the number of vessels ofosteoporotic and myeloma patients.

The percentage of surface area occupied bysinusoids was identical in plasma cell disorders,whatever the stage, in osteoporosis and RBM. It

Laroche et al.

90

was decreased by half in WaldenstroÈm's macro-globulinemia.

Trabecular bone volume was normal in MGUSand in stage I myeloma, and was moderatelydecreased in stage III myeloma. The decrease intrabecular bone volume was linked with increasedbone resorption and decreased osteoid parameters,while the thickness of cortical bone remained thesame.

In WaldenstroÈm's macroglobulinemia, no changein TBV or bone remodeling was observed.

Only patients with MGUS, stage I and IIImyeloma showed a TBV inversely correlated withthe number of arterioles (r=x0.35; p=0.05) andtheir eroded surfaces were moderately correlatedwith the number of arterioles: r=0.29; p=0.07).

No correlation was found between bone marrowcellularity and the number of capillaries (r=0.1,p=0.6) or the number of arterioles (r=0.2, p=0.3).The only correlations found related to: bone marrowcellularity and the surfaces occupied by sinusoids:r=x0.5, p=0.01; surfaces occupied by adipocytesand surfaces occupied by sinusoids: r=0.6, p=0.01.

There was no correlation between plasma cellin®ltration measured by bone marrow count and thenumber of vessels.

Discussion

We have shown that vascularity (arterioles andarterial capillaries) is increased in all stages ofplasma cell disorders and in WaldenstroÈm's macro-globulinemia (Fig. 1).

It seems acceptable to consider patients withosteoporosis as controls since Burkhardt showed in

1987 that the number of capillaries was identical inpatients with osteoporosis and in controls (8). Thesame author had, however, shown in another study(9) that in certain cases of osteoporosis withincreased bone resorption (hyperthyroidism orhyperparathyroidism), the number of arterial capil-laries increased. For this reason, we matchedpatients with osteoporosis and those with myelomaaccording to resorption surfaces, voluntarily select-ing cases of osteoporosis with increased resorption.

Vascularity of reactive bone marrows was higherthan that of osteoporosis but two-fold lower thanthat of MGUS or myeloma, although bone marrowcellularity was identical. Thus, in MGUS ormyeloma, the proliferation of plasma cells inducesan increase of vascularity, independent of bonemarrow cellularity.

Increased vascularity in plasma cell disordersinvolves only the arterial compartment (arteriolesand capillaries) and appears very early, as soon asthe gammopathy becomes symptomatic. It increasesmoderately with the stage of gravity of the diseasewhatever the marrow cellularity or the percentageof plasma cell in®ltration. Increased vascularity wasalso apparent in WaldenstroÈm's disease.

The percentage of surface occupied by sinusoidswas the same in MGUS, myeloma and osteoporo-sis; it was lower in WM and was inversely correlatedwith marrow cellularity.

Vacca et al. (10) also showed the existence of aneovascularization in myeloma. They did not applythe same methodology as we did for vesselquanti®cation. Indeed, in order to label the vessels,they used the anti-factor VIII murine monoclonalantibody which does not distinguish between thedifferent types of vessels. Consequently the surfaces

Table 1. Bone marrow cellularity and number of intra-osseous vessels in the various disorders

MGUS

(n=15)

Stage I

myeloma

(n=15)

Stage III

myeloma

(n=15)

WaldenstroÈm

(n=7)

RBM

(n=10)

OP

(n=10)

p

(Anova)

Hematopoiesis (%) 44.0t10 54.2t10 59.1t11 72.0t11 53.8t10 40.1t14 0.004

Arterioles/mm2 3.3t2 4.5t1.7 4.9t2 8.2t1.9 2.7t1 1.2t1.1 0.01

Capillaries/Mm2 21.6t8.1 22.0t5 26.0t7.2 22.3t10.4 12.6t1.2 4.4t1.8 0.001

Sinusoids/mm2 6.1t2.3 7.2t2.1 7.8t1.8 3.0t3.65 6.4t1.8 9.6t2.5 0.05

RBM, reactive bone marrow; OP, osteoporosis.

Table 2. Histomorphometric parameters in the various patient groups

MGUS Stage I myeloma Stage III myeloma WaldenstroÈm Osteoporosis

TBV 26.7t3.8 23.0t6 16.2t3.3 24.2t5 14.8t4.2

ES 5.0t1.5 5.6t3 8.0t4 4.2t0.8 5.8t2.3

OS 19.0t7 19.3t7 12.0t8.9 14.2t12 16.5t5.5

OV 4.7t2 4.8t2.3 2.5t2.8 3.2t2.4 3.0t2

TBV, trabecular bone volume; ES, eroded surfaces; OS, osteoid surfaces; OV, osteoid volume.

Increased vascularization in myeloma

91

occupied by the microvessels were larger in theactive MM than in the MM in relapse or in the non-active myeloma.

The data depend on the methodology used toquantify the vascular network because Rajkima et al.(11) related a hypervascularization in the myelomacompared with Hodgkin's disease and, at differencewith Vacca, these authors did not observe any down-regulation of the myeloma vascularization aftertreatment even on patients with a completeresponse.

Angiogenesis and microvascularization have fre-quently been studied in solid tumors and angiogen-esis appears to be indispensable to the developmentof these tumors, predicting their aggressivity;invasive breast cancers have high microvesseldensity (1). However, tumor cell proliferation andvessel proliferation appear to be independent of eachother (12). Vascularity is two-fold greater in high-grade than in low-grade prostate cancers (13). Invitro and in the animal, partial suppression ofneovascularization by angiostatic pharmacologicalagents decreases tumor growth (14).

Growth factors stimulate cancer angiogenesis:vascular endothelial growth factor (VEGF), basic®broblast growth factor (BFGF) and placentagrowth factor (PlGF) in renal carcinoma (15),epidermal growth-factor receptor (EGFR) inbreast cancer (16), transforming growth-factorbeta (TGF-beta), tumor angiogenic factor (TAF)and vascular permeability factor (VPF) in otherneoplasias (17±19). Prostaglandins, i.e. PGE1 andPGE2, and cytokines (IL1, IL8, TGFb, TNFa,GMCSF) could stimulate the secretion of angioge-netic factors by the macrophages: these cytokinesare responsible for the increase in angiogenesis inrheumatoid synovium and could be involved in theangiogenesis of the reactive bone marrows (20).

Could IL1 and TNF-alpha, which are classicallyoverproduced in myeloma (21), in¯uence angiogen-

esis? Can we implicate BFGF, which can be secretedby myelomatous cells (22), or VEGF, which is highin myeloma and in the POEMS syndrome inparticular (23)?

The number of arterioles is correlated in anegative way with the bone trabecular volume andin a positive way with the eroded surfaces. We knowthat the osteoclastic hyper-resorption of myelomadepends at least on IL6 . Does this cytokine act as anangiogenetic factor? Could some growth factorssuch as VEGF, for which receptors have beencharacterized on osteoblasts, or matrix metallopro-teinases be implicated in bone remodelling?

References

1. WEIDNER N, SEMPLE JP, WELCH WR, FOLKMAN J. Tumorangiogenesis and metastasis: correlation in invasive breastcarcinoma. N Engl J Med 1991;324:1±8.

2. FOLKMAN J. What is the evidence that tumors areangiogenesis dependent? J Natl Cancer Inst 1990;82:4±6.

3. WEIDNER J. Tumor angiogenesis: basic concepts and reviewof current applications in tumor prognostication. SeminDiag Pathol 1994;10:302±313.

4. KYLE RA. Monoclonal gammopathy of undeterminedsigni®cance: natural history in 247 cases. Am J Med1978;64:8714±8726.

5. DURIE BGM, SALMON SE. A clinical staging system formultiple myeloma. Cancer 1975;36:842±854.

6. PARFITT AM, DREZNER MK, GLORIEUX F, et al. Bonehistomorphometry: standardization of nomenclature, sym-bols and units. J Bone Miner Res 1987;2:595±610.

7. AZIZA J, MAZEROLLES C, SELVES J, et al. Comparison of thereactivities of monoclonal antibodies QBENd10 (CD 34)and BNH9 in vascular tumors. Appl Immunohistochem1993;1:51±57.

8. BURKHARDT R, DEMMLER K, BARTL R, et al. Les rapportsmorphologiques entre la circulation de la moelle et celle del'os dans les conditions normales et pathologiques. In:Circulation Osseuse, Paris: Armor Montagu, 1997:221±247.

9. BURKHARDT R, KETTNER G, BOHM W, et al. Changes intrabecular bone, hematopoiesis and bone marrow vessels inaplastic anemia, primary osteoporosis and old age: acomparative histomorphometric study. Bone1987;8:157±164.

10. VACCA A, RIBATTI D, PRESTA M, et al. Bone marrowneovascularization, plasma cell angiogenic potential andmatrix MMP2 secretion parallel progression of humanmultiple myeloma. Blood 1999;93:3064±3073.

11. RAJKUMAR SV, FONSECA R, WITZIG TE, GERTZ MA, GREIPP

PR. Bone marrow angiogenesis in patients achevingcomplete response after stem cell transplantation formultiple myeloma. Leukemia 1999;13:469±472.

12. VARTANIAN RK, WEIDNER N. Correlation of intratumoralendothelial cell proliferation with microvessel density andtumor cell proliferation in breast carcinoma. Am J Pathol1994;6:1188±1194.

13. FREGENE TA, KHANUJA PS, NOTO AC, et al. Tumor-associated angiogenesis in prostate cancer. AnticancerRes 1993;13:2377±2382.

14. WEINSTAT-SASLOW D, STEEG PS. Angiogenesis and coloniza-tion in the tumor metastatic process: basic and appliedadvances. Faseb J 1994;8:401±407.

15. TAKAHASHI A, SASAKI H, KIM SJ, et al. Markedly increasedamounts of messenger RNAs for vascular endothelial

Fig. 1. Iliac crest bone marrow biopsy. Goldner stain:numerous arterial capillaries within a malignantproliferation of plasma cells (stage III myeloma).

Laroche et al.

92

growth factor in renal cell carcinoma associated withangiogenesis. Cancer Res 1994;54:4233±4237.

16. FOX SB, LEEK RD, SMITH K, HOLLYER J, GREENALL M,HARRIS AL. Tumor angiogenesis in node-negative breastcarcinomas: relationship with epidermal growth factorreceptor, estrogen receptor, and survival. Breast CancerRes Treat 1994;29:109±116.

17. FAN PT. Angiosuppressive therapy for cancer. TrendsPharmacol Sci 1994;15:33±36.

18. UEKI N, NAKAZATO M, OHKAWA T, et al. Excessiveproduction of transforming growth factor beta 1 can playan important role in the development of tumorigenesis.Biochim Biophys Acta 1992;1137:189±196.

19. KONDO S, ASANO M, SUZUKI H. Signi®cance of vascularendothelial growth factor/vascular permeability factor

for solid tumor growth, and its inhibition by theantibody. Biochem Biophys Res Commun1993;194:1234±1241.

20. FOLKMAN J. Angiogenesis in cancer, vascular, rheumatoidand other disease. Nature Med 1995;1:27±31.

21. BATAILLE R, CHAPPARD D, KLEIN B. Mechanisms of bonelesions in multiple myeloma. Hematol Oncol1992;6:285±295.

22. FILLOLA G, LAFFARGUE M, BOUSQUET R, et al. Basic®broblast growth factor is produced by myeloma celllines. Br J Haematol (in press).

23. SOUBRIER M, DUBOST JJ, AUTHIER FJ, et al. Les facteurs decroissance dans le syndrome POEMS: mise en eÂvidenced'une augmentation importante du vascular endothelialgrowth factor. Rev Rhum 1996;14:798.

Increased vascularization in myeloma

93