Document ZDpyVXQ90ajgjn75wzwwyGKY

#? ,iI > - j Id 664 THE NEW ENGLAND JOURNAL OF MEDICINE 19. Lee K-s, Ferguson RM. Corpuz M, Gartner LM. Maternal age and incidence of low birth weight at term.Am J Obstet Gynecol 1988; 15834-9. 20. Kiely JL, Paneth N, Susser M. An assessmentof the effectsof maternal age and parity in different components of perinatal monality. Am J Epidemiol 1986; 123:444-54. 21. Naeye RL. Maternal age, obstetric complications, and the outcome of pregnancy. Obstet Gynecol 1983; 61:210-6. 22. Hardy JB. Mellits ED. Relationshipof low birth weight to maternal characteristics of age, parity, education and body size. In: Reed DM, Stanley FJ, eds. The epidemiology of prematurity: proceedings of the epidemiology workshop held at NICHDINIH, Bethesda, Md., November 8-9, 1976. Baltimore: Urban & Schwarzenberg. 1977:105-17. 23. Kramer MS. Determinantsof low binh weight: methodological assfiur and meta-analvsis. Bull WHO 1987: 65:663-737. mp ~~ ~ ~~~~~ I - ~ - ~ ~~~ ~ 24. Berkowitz GS. Clinical and obstetric risk factors for preterm delivery. Sinai J Med (NY)1985; 52239-47. 25. Harlap S, Shiono PH, Ramcharan S. A life table of spontanews and the effectsof age, parity. and Other variables. In: Porter IH, Hook U, eds. Human embryonic and fetal death. New York: Academic %: 1980 145-58. 26. Hook EB. Obstet G~ Rates of chromosome IECO19~81; 58282-5. abnormalities at different matem& aga. 27. Bulmer MG. The effect of parental age, parity and duration of wage the twinning rate. Ann Hum Genet 1959; 23:454-8. MARKERS OF MULTIPLE HEMATOPOIETIC-CELL LINEAGES I N MULTIPLE MYELOMA JOSHUA EPSTEIND, .Sc., HUIQINGXIAO, M.D., AND XIAO-YAN HE, M.D. Abstract Multiple myeloma is considered a cancer of mature plasma cells. Recent studies, however, suggest the possible involvement of early B cells and the ex- pression of myelomonocytic antigens by myeloma cells. Using flow cytometry, we searched for evidence of the expression of genes specific for different hematopoietic lineages by tumor cells in bone marrow aspirates from 27 patients with aneuploid multiple myeloma. In addition to features characteristic of myeloma cells, we found evidence of the frequent expression by myeloma tumor cells of the pre-Ball antigen CALLA (common acute lympho- cytic leukemia antigen) (in specimens from 58 percent of patients) and of megakaryocytic (88 percent), m y e l o m ocytic (65 percent), and erythroid (39 percent) surfam markers. The proportion of tumor cells expressing the &if. ferent markers varied among patients, from 2 to 100 per- cent of recognizable tumor cells. We conclude that cells of multiple lineages are in- -volved in myeloma a finding that is consistent with the hypothesis that there is a common primary neoplastic lesion for all hematologic cancers. (N Engl J Med 1990;322: 664-8.) DI IFFERENTIATION of hematopoietic progeni- cell lines and supports the hypothesis that there is a tor cells into lineages is accompanied by the acti- common primary neoplastic lesion in all hematologic vation of genes that regulate the expression of surface antigens specific to each cell line and each stage of differentiation. These antigens, known as surface markers, are used in the diagnosis and classification of hematologic neoplasias because they indicate the lineage and stage of differentiation of the tumor cells.' Multiple myeloma is a B-cell cancer, typically associated with mature plasma-cell morphology and func- cancers. To determine how frequently and to what extent different hematopoietic lineages are involved in plasma-cell myeloma, we measured the expression of lineage-associated genes by tumor cells in bone marrow aspirates from patients with myeloma. We used dual-parameter flow cytometry to identify the expression of megakaryocytic, myelomonocytic, and erythroid markers by aneuploid myeloma cells. tion. There is, however, increasing evidence of the involv6ment of early B lymphocytes in METHODS Among the evidence for the presence of neoplastic pre-B cells in myeloma is the expression of the preB-cell antigen CALLA (common acute lymphocytic leukemia antigen), often without cytoplasmic immunoglobulins, in the bone marrow of about 50 percent of patients with m y e l ~ m a , ~th-e~ presence of pre-B cells in peripheral blood,5 and the differentiation of CALLA-bearing lymphocytes to monoclonal plasma cells with concordant isotypes.6 More recent studies have demonstrated the presence of myeloid antigens on myeloma cells and found them to be associated with a poor progno~isT.~he presence of tumor cells expressing myeloid antigens in this B-cell neoplasia suggests the involvement of other hematopoietic Patients We studied 27 patients with a diagnosis of DNA-aneuploid myeloma. All the patients gave informed consent, and the study was approved by the human studies committee of the University of Arkansas institutional review board. Twelve patients were studied at the time of diagnosis, before the initiation of treatment, and six at relapse. Seven patients with disease that was unresponsive to therapy were studied after the end of treatment, and two who had responded to treatment were studied between courses of chemotherapy. Bone marrow aspirates treated with heparin were subjected to Ficoll-Hypaque density separation (specific gravity, 1.077 g per cubic centimeter). Light-density mononuclear cells collected from the plasma-Ficoll-Hypaque interface were examined for ploidy and monoclonal cytoplasmic immunoglobulin by DNA-cytoplasmic immunoglobulin Row cytometry as previously described." Because there were insufficient quantities of cells to perform all the cell assays on specimens from all patients, the order of laboratory studies was randomly assigned so a s to minimize inadvertent bias. From the Division of HematologylOncology,Universityof Arkansas for Medical Science, Arkansas Cancer Resuurh Center, Little Rock. Address reprint quests to Dr. Epstein at Univelsity of Arkansas'Medical School, 4301 W. Markham St., Slot 508. Little Rock, AR 72205. Supported in part by grants (CA 37161 and CA 28771) from the National Institutes of Health. Reagents T h e expression of myeloid, megakaryocytic, erythroid, and 1ympi:oid antigens by myeloma tumor cells was probed with monoclonal antibodies. Table I lists the antibodies used, the antigen and -~ bl. 322 No. 10 M U L T I P L E HEMATOPOIETIC-CELL LINEAGES IN MYELOMA EPSTEIN ET AL. 665 aMe 1. Probesfor the Expressionof Lineage-AssociatedGenes. KmoDY ANTKiEN LINEAGE SouKE' ASSAY CDlO Lymphoid. Coultcr Direa myeloid t1 GPIIb/Illa Megakaryocytic Coultcr Dircct bl CDllb Myelomonocytic, Coulter T-cell subst Dirat lP CD33 Myeloid 3 - 1 26-kd Antigen Pbma cells b39 Glycophorin A Erythroid Coulter Coultcr Griffin lndirat Indirat Indirect mdenncr Coultm I m u d o g y , Hulcrh, Hi.;IIKantibody to glycophonn A was a trom h 1. Griffinof UK Datu-Fubcr Canox Instl~cB. eam lineage they recognized, their source, and whether direct or indirect immunofluorescence assays were employed. All the monoclonal antibodies were used under the conditions recommended by the i dsuppliers. Isotype controls were obtained from the same sources and under identical conditions. studies were carried out with use of dual-parameter flow . Initially, flow cytometry was used to determine the phe- aneuploid cells. Aliquots containing 0.5to 1.OX 106 mono11s were reacted with fluorescein isothiocyanate-conjugatonoclonal antibodies, washed, and fixed in 70 percent ice-cold anol for 1 to 14 hours. T h e cells were then washed, treated with ribonuclease, counterstained with propidium iodide, and analyzed n a FACScan flow cytometer (Becton Dickinson) as previously described? When fluorescein isothiocyanattconjugatedantibody was not available, a second antibody, fluorescein isothiocyanate- conjugated rabbit antimouse immunoglobulin (Dako), was used in an indirect assay. To identify coexpression of surface antigens and cytoplasmic light chain, cells were reacted with monoclonal antibodies and futed as described above, then reacted with phycoerythrin-conjugated goat antihuman light-chain F(ab'), fragments (Tago). To analyze the coexpression of nonlymphoid and plasmacell-specific antigens, cells were reacted with the purified monoclonal antibody PCA-I and then with phycoerythrin-conjugated goat antimouse immunoglobulin, blocked with normal mouse serum, and finally reacted with the second fluorescein isothiocya- nate-conjugated monoclonal antibody. For each sample, isotype- matched and, when indicated, opposite-light-chain controls were run simultaneously. In each case, overlap of emission spectra was compensated for electronically with use of the proper control samples. RESULTS The results of our analysis of the expression of lymphoid, megakaryocytic, erythroid, and myelomonocytic antigens by DNA-aneuploid myeloma tumor cells are illustrated in Figure 1. In addition to reacting with the anti-CD10 antibody 5-5, cells in the hyperdiploid compartment also expressed the glycoprotein IIb-IIIa complex (GPIIb/IIIa), CDI lb, and CD33. The reactivity with anti-glycophorin A was questionable. The different antigens were expressed independently, as illustrated in Figure 1B, where hypodiploid myeloma cells expressing GPIIb/IIIa showed no reactivity with the anti-CD1l b reagent Mol. Table 2 summarizes the results of phenotype P DNA Content D DNAconMn D AB b t Figure 1. Multilineage Antigen Expression in Aneuploid Myeloma. PanelAshows,tumor cells, identifiedbytheir hyperdiploidDNA content, reactedpositivelywith antibodiesagainstCALLA (11 percent Positive), GPllb/llla (10 percent), CD33 (57 percent),and CDllb (13 percent). Reactivitywith antibody to glycophorinA was inconcluO1 *e. The results for isotype-matched controls are shown in the top row. The abscissa indicates the degree of red fluorescence of I PWidium iodide as an indicatorof DNA content. The ordinate indicates the frequency of green fluorescenceof fluorescein isothiocya- f fWeasan indicatorof the phenotype. Hypodiploidtumor cells (Panel6).althoughthey expressedthe megakaryocyticantigenGPIIWllla, i did not react with the anti-CD11b antibody Mol. I 666 I THE NEW ENGLAND JOURNAL OF MEDICINE March 8, I 1 analyses of cells from all 27 patients. Expression of CDlO was recorded in samples from 58 percent, GPIIb/IIIa in 88 percent, CDI Ib in 65 percent, and glycophorin A in 39 percent of the patients studied. The proportion of myeloma tumor cells expressing the different markers varied from patient to patient; CALLA was detected in a mean (+SD) of 15210 percent (median, 9 percent; range, 5 to 31 percent), GPIIb/IIIa in 53+37 percent (median, 32 percent; range, 7 to 100 percent), CDIIb in 38+34 percent (median, 20 percent; range, 5 to 100 percent), and I glycophorin A in 27+36 percent (median, 7 percent; range, 2 to 98 percent). In addition to these antigens, I CD33 was expressed by tumor cells from all four pa- tients examined. I Since in many cases only a fraction of the aneuploid I cells reacted with the various antibodies, it was important to identift the coexpression of characteristi- cally myeloma-associated and nonlymphoid markers. I Figure 2 shows the results of such a study in a patient with marked marrow plasmacytosis (88 per- I Table 2. Expression of Lineage-AssociatedGenes by Myeloma I Tumor Cells. DNAcontent A PATIENT CATEGIJRY CDlO ANTIGEN GPUbllna CDllb GLYCOPUOUN A M.expressing Onripenlno.s i d e d I Studied at diagnosis 1/12 11/12 7/12 3/10 Studied at relapse 2/4 315 a5 315 Resistant* 316 7/7 6/7 26 Othert 2 2 2n 2/2 10 Total 14/24 23/26 17/26 9/23 *Patients with d i e th.1 was unresponsiveto hutment. sndicd after tMmYnt mdcd. tPatients who rcspondcdIO I l r r w n t who wuc studid b*rveen nnusesof chcmomcmpY. cent of plasma cells) whose hyperdiploid tumor cells expressed both PCA- 1 and GPIIb/IIIa. Although PCA-I was expressed by 83 percent and GPIIb/IIIa by 20 percent of the cells (Fig. 2A), both antigens were present in 16 percent of all cells (Fig. 2B). Similarly, Figure 3 shows the coexpression of cytoplasmic kappa light chain and megakaryocytic, myelomonocytic, and erythroid antigens in bone marrow cells from a patient with substantial marrow plasmacytosis (90 percent of cells) and a high proportion of cells (88 percent) containing cytoplasmic immunoglobulin-kappa. Cytoplasmic immunoglobulin-kappa was coexpressed with GPIIb/IIIa, glycophorin A, and CDI Ib by 10, 22, and 16 percent of cells, respectively. DISCUSSION To evaluate the possible involvement of several hematologic cell lineages in this B-cell neoplasia, we analyzed the expression by myeloma tumor cells of genes that are specific for hematopoietic cell lines and for particular stages of differentiation. The expression of such markers has been a useful tool in the classification of hematologic cancers. The close association between the unexpected expression of lineage and differentiation markers and the clinical features of some Log RuontscenceIntensity (FITC) B Figure 2. Coexpression of Plasma-Cell Antigen (PCA-1) and Megakaryocytic Antigen (GPllb/llla) in a Patient with Myeloma In the correlated DNA-phenotype analysis (Panel A), hyper- diploid cells expressed GPllWllla and PCA-1. Panel B show8 contour plots representing the densitv distribution of cells ma- both antigens. Wrthin the lower right-hand section, PCA-l-pmb ' tive, GPIIWllla-negative cells are represented in the upper leB * ' quadrant, PCA-l-negative. GPIIWllla-positive cells in the kwa right quadrant, and PCA-l-pcsitive. GPllWllla-positive Celts the upper right quadrant. The lower left quadrant shows negativefor both PCA-1 and GPIIWllla. The upper left-hand tion shows cells reacted with the isotype-matched munoglobulinsfor bothantibodies.The upper right-handand er left-handsections show the resultsfor individualcontrols. Edh' antibodies were IgM. PE denotes phycoerythrin. and FITC fluorescein isothiianate. ~ MULTIPLE HEMATOPOIETIC-CELL LINEAGES IN MYELOMA - EPSTEIN ET AL. 667 proportions of tumor cells found to express the different antigens. In a similar study of patients with myeloid leukemia, the associ- ation of lineage-specific antigens with specific cytogenetic anomalies was used to demonstrate the neo- plastic involvement of morphologi- cally normal lineages, suggesting that a multipotent hematopoietic progenitor cell was involved in the disease." In contrast, the coexpres- control GPlMa Gtymplmnn A CDI 1b sion of typical myeloma-cell fea- tures and nonlymphoid antigens Log Fluorescence Intensity (FITC) might be interpreted as ruling out Figure 3. Density Distributionof Cells from a Patientwith a Kappa-ExpressingDiploid 1 MyelomaAnalyzed for the Expression of Monomorphic CytoplasmicImmunoglobulin and Myelomonocytic, Megakaryocytic,and Erythroid Antigens. the presence of megakaryocytes and myeloid and erythroid cells in the DNA-aneuploid cell compart- Thetop and bottomrowsrepresent cells reactedwith anti-kappa and anti-lambda light chains, respectively. Within each panel, the upper left quadrant contains cells that matted with the ordinate antibody, the lower right quadrantcells that reactedwith the ab8dssa antibody, and the upperright quadrant cellsthat reactedwith bothantibodies and hence coexpress the two antigens. Cells that reacted with neither antibody are Shown in the lower left quadrant. PE denotes phycoetythrin, and FITC fluorescein isothiocyanate. ment; this suggestion is compatible with the apparently normal hematopoiesis observed in patients with myeloma. Rather, it suggests that the primary neoplastic lesion in plasma-cell myeloma occurs in an early hematopoietic progenitor ca- hematologic tumors indicates that at least in some pable of differentiation into the various lineages. This htances, the expression of these genes may have a hypothesis is supported by a previous report of re- role in the evolution of disease and should not be dis- arrangements of T-cell receptor gamma-chain genes missed as an aberration, coincidental to the neoplasia and invariant chain expression in It is or reflecting a loss of regulatory control associated still possible that, as observed in the study of myeloid with malignant transformation. The expression of leukemia,'' different lineages may bear yet unidenti- PreB-cell and myeloid antigens by myeloma cells fied genetic abnormalities associated with myeloma. has prognostic implication^,^,^,^ and the expression of Taken together, all the studies reporting lineage and CALLA led to the understanding of pre-B-cell in- differentiation infidelity and multilineage involvement volvement in a disease typically seen as a tumor of in hematologic cancers could imply the existence of a mature B ~ e l l sT.h~e e~xp~re~ssi~on of megakaryocytic common neoplastic progenitor for all hematologic tu- and erythroid genes, in addition to lymphoid and mye- mors, with further events determining the phenotypic loid genes, by myeloma cells is now being reported presentation of the disease. by investigators using DNA ploidy, cytoplasmic im- The involvement of multiple lineages provides clues munoglobulin content, and the presence of a plasma- for mapping the sequence of the transforming events cell-specific antigen as three independent markers of in terms of the development of the hematopoietic sys- myeloma tumor cells. tem. The high incidence of multilineage involvement Although it was not the purpose of this study to in myeloma as well as other hematologic cancers could elucidate the prognostic implications of our findings, validate the sequential model of lineage commitment the presence of the different markers appears to be during hematopoietic development15 by suggesting unrelated to the disease stage. The observed frequen- CY of "lineage infidelity" reflected by the expression Of multilineage-associated genes by myeloma cells in this study (39 to 81 percent) suggests that this phenomenon is common. The absence of multilineage mvolvement in some patients probably reflects the low density of antigen presentation - below the lev- el that can be detected by the available methods - that the neoplastic cells express genes representing the "stations" they have traversed during differentiation and lineage commitment. The coexpression of various lineage-associated genes by tumor cells could suggest that cells with less stringent differentiation, which are likely to be present in small numbers and for short periods during normal hematopoietic development, are immortalized by the neoplastic transformation. on the surface of myeloma plasma cells. This phe- Alternatively, cells bearing neoplastic lesions could nomenon is best exemplified by the My9 antibody; have lost the linkage between lineage commitment reactivity was detected when the fluorescein iso- and the expression of surface molecules that deter- aqanate-conjugated reagent was used, but mye- mine microenvironmental interactions (e.g., homing), loma cells expressing CD33 were readily detected in resulting in the many manifestations of each hemato- four bone marrow samples studied with the more logic cancer. f e m i h indirect method. Differences in the sensitiv- We arc indebted to Dr, Bart Barlogic for his interest In and 'fy of the methods may also account for the varying support of this work. '\ - - L 71668 T H E NEW ENGLAND JOURNAL OF MEDICINE March 8, 1990 REFERENCES 1. Foon KA, Todd RF Ill. Immunologic classification of leukemia and lymphoma. Blood 1986: 6&1-31. 2. Eptein J, Barlogie B, Katzmann F. Alexanian R. Phenotypicheterogeneity in aneuploid multiple mysloma indicates pre-B cell involvement. Blood 1988; 713861-5. 3. Grogan TM, Dune BG, Lomen C. et al. Delineation of a novel pre-B cell comparumnt in plasma cell myeloma: immunochemical, immunophenotypic, genotypic, cytologic, cell culture, and kinetic features. Blood 1987; 70932-42. 4. Dune GM,Grogan TM. CALLA-positive myeloma: an aggressive subtype with poor survival. Blood 1985;6229-32. 5. Pilarski LM, Mant MI, Ruether BA. F're-B cells in peripheral blood of multiple myeloma patients. Blood 1985; 66:416-22. 6. Caligaris-CappioF, Bergui L, Tesio L, et a). Identification of malignant plasma cell precursors in the bone manuw of multiple myeloma. J Clin Invest 1985; 761243-51. 7. Ruiz-Arguelles GJ, Katzmann JA, Greipp PR, Gonchoroff NJ, Garton JP, Kyle RA. Multiple myeloma: circulating lymphocytes that express plasma cell antigens. Blood 1984;64:352-6. 18. Mellstcdt H,Hammantrom S.HolmG.Monoclonal lymphocytepopulptim in human plasm cell myeloma. C h h p l m m U n 0 l 1974; 17371-84. I 19. Grogan TM.h r i e BG, Spier CM, Richter L, Vela E. Myelo~nowcylic antigen positive multiple myeloma. Blood 1989; 73:763-9. '10. BarlogicB, Alexanian R, Pershwse M. SmallwoodL, Smith L. Cy~oplrc. 1mic immunoglobulin content in multipk myeloma. I Clin Invest 198% 76765-9. ,1I . Keiniinen M. Griffin ID. BloomfieldCD, Machnicki J, & la chapeile A, Clonal chromodomal abnormalities showing multiplecell-lineage i n v d w Iment in acute myeloid leukemia. N Engl J Med 1988; 3181153-8. 12. Lee M, SelvanayagamP,Alexanian R. Stass S.BarlogieB. - g e m (REARR)of T-cell receptor(TcR) y chain gene in multiplemyeloma (MM). Roc Am As= Cancer Res 1987; 28:178. abstract. 13. Nami F,SelvanayagamP. Barlogie B. SaundersOF. HLA-associated inva. i iant chain gene expression in human B cell neoplasia. Leukemia 1987; 1504-7. 14. lBoamrlaog-ienBew, EbpiostteoignicJa,l Selvanayagam P, Alexanian R. Plasma insights and advances in therapy. Blood cell mye1989 73: 865-79. 15. Brown G , Bunce CM, Howie AI. Lord JM. Stochastic or ordered linuge ; ! ; . commitment during hemopoiesis?Leukemia 1987; 1:150-3. ,I 1 Massachusetts Medical Society Registry on Continuing Medical Education ~ To obtain informationon continuing medical education courses in the New England area, call between 9:00 a.m. and 12:OO noon, Monday through Friday, (617)893-4610 or in Massachusetts 1-800-322-2303, ext. 1342. If writing, direct correspondence to: Program Registrar, Massachusetts Medical Society, 1440 Main St., Waltham, MA 02154-1649.The booklet is free to MMS members, $5.00 for nonmembers.