Fluorescence in situ hybridization analysis of aneuploidization patterns in monoclonal gammopathy of undetermined significance versus multiple myeloma and plasma cell leukemia
Article first published online: 17 JAN 2003
Copyright © 2003 American Cancer Society
Volume 97, Issue 3, pages 601–609, 1 February 2003
How to Cite
Ana, R., Dolores, T. M., Luz, S. M., de Andrés Martín, P., Marta, M. A., José, H., Jesús, M. M., Javier, F.-C., María, S. J., Aglae, B., Fernando, S. M. J. and Alberto, O. (2003), Fluorescence in situ hybridization analysis of aneuploidization patterns in monoclonal gammopathy of undetermined significance versus multiple myeloma and plasma cell leukemia. Cancer, 97: 601–609. doi: 10.1002/cncr.11100
- Issue published online: 17 JAN 2003
- Article first published online: 17 JAN 2003
- Manuscript Accepted: 9 SEP 2002
- Manuscript Revised: 5 SEP 2002
- Manuscript Received: 9 JUL 2002
- monoclonal gammopathy of undetermined significance (MGUS);
- interphase fluorescence in situ hybridization (FISH);
- chromosome numerical abnormalities;
- plasma cell clones
Monoclonal gammopathy of undetermined significance (MGUS) is a clonal plasma cell (PC) disorder usually characterized by a benign clinical course. However, in approximately 25% of patients, the disorder has been found to evolve into a multiple myeloma (MM). The mechanism leading to the evolution of MGUS remains unknown. The aim of the current study was, first, to assess by interphase fluorescence in situ hybridization (FISH) the incidence of numerical abnormalities of chromosomes 6, 9, 13, and 17 in MGUS patients and to compare it with that found in MM and PC leukemia (PCL) patients and, second, to explore the potential heterogeneity of the pathologic PC in MGUS as a way to identify unique cytogenetic patterns different from those frequently observed in MM and PCL.
Numerical abnormalities of chromosomes 6, 9, 13, and 17 were investigated by dual- and triple-color FISH in bone marrow PC from 208 patients corresponding to MGUS (n = 30), MM (n = 158), and PCL (n = 20) cases. In MGUS and MM patients with < 10% PC, both normal and phenotypically aberrant PC were discriminated by multiparameter flow cytometry, the latter subset being specifically sorted for FISH analysis with a purity of 93% ± 6%.
Overall, 57% of the MGUS patients displayed abnormalities for at least 1 of the 4 chromosomes analyzed compared with 75% of both MM and PCL cases. The most common single chromosome abnormalities detected in MGUS were gains of chromosomes 9 (23%) and/or 6 (21%) and loss of chromosomes 13 (21%) and/or 17 (17%). Compared with MM patients, MGUS patients were found to have both a lower incidence of gains of chromosome 9 (23% vs. 54%, P = 0.002) and monosomy 13/13q- deletions (21% vs. 38%, P = 0.07); with respect to PCL cases, MGUS patients were found to have a lower incidence of monosomy 13/13q- deletions (21% vs. 75%, P < 0.001) together with a slightly higher frequency of gains of both chromosomes 6 (21% vs. 0%, P = 0.05) and 9 (23% vs. 7%, P = 0.1). The simultaneous use of two or three different chromosome probes showed that within the purified compartment of phenotypically aberrant PC from most MGUS patients (67%), more than 1 PC clone could be identified. In contrast, the incidence of 2 or more PC clones was much lower in MM (19%, P < 0.001) and PCL (15%, P = 0.003). Interestingly, although some FISH patterns were shared by both groups of diseases (i.e., monosomy 13/13q- deletions alone, gains of chromosome 9 alone or together with trisomy 6), others were found almost exclusively in either MGUS (i.e., a clone with monosomy 6 and/or 17 together with nuclei displaying a normal chromosome number) or in MM (i.e., monosomy 13/13q- deletions together with gains of chromosome 6 and/or 9).
In summary, the results of the current study showed that MGUS patients displayed a high incidence of numerical alterations, which are usually associated with the presence of more than one tumor cell clone. It is interesting to note that the cytogenetic patterns observed in the aneuploid PC clones from MGUS patients were frequently different from those observed in both MM and PCL. Cancer 2003;97:601–9. © 2003 American Cancer Society.
Monoclonal gammopathy of undetermined significance (MGUS) is the most frequent clonal plasma cell (PC) disorder with an overall incidence of 62%, which corresponds to 1% of the population age > 50 years and 3% of the population age > 70 years.1, 2 Although in the majority of cases MGUS displays a benign clinical course, in approximately 25% of the patients the disease has been found to evolve into a multiple myeloma (MM) or another malignant B-cell lymphoproliferative disorder,2, 3 and up to 33% of newly diagnosed MM patients may have a previous history of MGUS.4 The mechanisms leading to the evolution from MGUS to MM are not fully understood,5 and currently no single factor has been identified to predict when an MGUS disorder will evolve into a malignant condition.2, 6 Moreover, differential diagnosis between MGUS and MM is based on the combination of several parameters, and regular follow-up of the M-component is frequently required.
In recent years, advances in the knowledge of the most frequent chromosome abnormalities present in MM7–9 and in plasma cell leukemia (PCL)10, 11 patients have been obtained due to a large extent to the systematic use of interphase fluorescence in situ hybridization (FISH) techniques.9 By contrast, in MGUS, information regarding the cytogenetic abnormalities present in clonal PC remains scant due to both their low numbers and proliferative index. Interestingly, interphase FISH studies have shown that chromosomal abnormalities frequently detected in MM and PCL, including those associated with a worse outcome (i.e., 13/13q-), also can be found in clonal PC in MGUS.12–17 Despite this, and to the best of our knowledge, no information has been provided on the cytogenetic patterns present in MGUS as established on the basis of specific aneuploidization pathways due to genetic chromosomal instability compared with those observed in MM and PCL. In addition, it should be noted that although clonal heterogeneity has been suspected14 by using cytoplasmic immunoglobulin (cIg) light chain (cIgκ+ or cIgλ+) plus interphase FISH,18–20 this technique does not allow the accurate assessment of intratumoral genetic heterogeneity in MGUS because variable proportions of normal residual PC may be included among the cIg light chain-restricted PC compartment.
The aim of the current study was to assess by FISH the incidence of numerical abnormalities of chromosomes 6, 9, 13, and 17 in MGUS patients and to compare it with that observed in both MM and PCL. We specifically selected these four chromosomes because they are frequently involved in MM, where they allow for the identification of almost every case carrying numerical chromosomal abnormalities.7 Our final goal was to explore the potential heterogeneity of the pathologic PC clone in MGUS, as a way to identify aneuploid cytogenetic patterns that are either similar to those frequently observed in MM and PCL or that could be of help to predict evolution from MGUS into a malignant clonal PC condition.
MATERIALS AND METHODS
Patients and Samples
A total of 208 newly diagnosed patients with monoclonal gammopathies including a core group of 30 MGUS and a reference control group of 158 MM and 20 PCL were included in the study. Diagnosis of MGUS was based on the following criteria: < 30g/L of serum M-component or a small amount of urine light chain protein excretion; < 10% bone marrow (BM) PC in the absence of lytic bone lesions on radiography; anemia; hypercalcemia; and an impaired renal function. Patients with MM were diagnosed according to the criteria of the Chronic Leukemia-Myeloma Task Force,21 and the diagnosis of PCL required > 2 × 109/L blood PC.22 The mean age of MGUS patients (17 males and 13 females) was 68 ± 13 years (range, 29–88 years). The reference group of MM (79 males and 79 females) and PCL (9 males and 11 females) cases had mean ages at diagnosis of 66 ± 11 (range, 40 to 90 years) and 66 ± 10 years (range, 45–81 years), respectively.
Within the MGUS patients, the M-component was IgG in 62% of cases and IgA in 38% of cases. The percentage of PC in BM was constantly < 10% (0.62–8.3%). At the end of this study, all except one patient (who died from causes unrelated to MGUS) remained alive and none showed progression in to MM; median follow-up was 24 ± 3.71 months (range, 18–31 months).
In all cases, EDTAK3-anticoagulated PBS-diluted (1/1; volume/volume [vol/vol]) BM samples obtained for diagnostic purposes were used for both immunophenotypic and FISH studies.
Immunophenotypic Studies and Sorting of BM PC
Immunophenotypic analysis of BM PC was performed as previously described.23, 24 The following combination of monoclonal antibodies in triple-stainings were systematically used: CD38/CD56/CD19; CD20/CD117/CD38; CD38/CD28/CD138; CD38/CD33/CD45. Analysis of reactivity of these monoclonal antibodies was performed on a FACScalibur flow cytometer (Becton Dickinson Biosciences, San Jose, CA) using the CellQuest software (Becton Dickinson Biosciences). In each case, results were analyzed for at least 103 BM PC using the PAINT-A-GATE software program. Identification of BM PC was based on their unique reactivity for the CD38hi and CD138+ antigens. Discrimination between normal and pathologic BM PC was based on the expression of aberrant phenotypes previously described for the above listed combinations of monoclonal antibodies.24
In all MGUS cases, as well as in MM patients with < 10% PC, immunophenotypically aberrant PC were sorted using a FACSVantage flow cytometer (Becton Dickinson Biosciences) equipped with 2 laser lines and 5 fluorescence detectors. Polygonal sorting regions were established on the basis of the PC′s light scatter properties, CD38hi and CD138+ expression, and aberrant antigen expression (CD56++, CD19−, CD28++, CD117++, and/or CD33+). Reanalysis of the sorted immunophenotypically aberrant PC showed a purity of 93% ± 6% (median, 93%; range, 76–99%).
Sorted cells were placed on up to eight different slides (a total of 5000 immunophenotypically aberrant PC/slide). Then the slides containing the sorted PC were fixed in 3:1 methanol/acetic (vol/vol) for 15 minutes at room temperature and stored at −20°C until analyzed by FISH.
FISH studies were performed in all cases on slides containing PC fixed in Carnoys solution and stored at −20°C. Numeric abnormalities for chromosomes 6, 9, 13, and 17 were systematically studied according to previously described methods.7 The following panel of DNA probes purchased from Vysis Inc. (Downers Grove, IL) were used in double and triple stainings: chromosome 6, CEP 6 DNA probe conjugated with spectrum orange (SO) and/or spectrum green (SG), which hybridizes to the centromeric region of chromosome 6 (6p11.1–q11.1); chromosome 9, CEP 9 DNA probe specific for the centromeric region of chromosome 9 (9p11–q11) labeled with SO and/or spectrum aqua (SA); chromosome 13, LSI 13 DNA probe conjugated with SO, which contains a DNA sequence specific for the Rb1 gene locus at the 13q14 chromosome region; and chromosome 17, CEP 17 DNA probe labeled with SG, which hybridizes to the centromeric region of chromosome 17 (17p11.1–q11.1).
Hybridization spots were evaluated under an Olympus BX60 fluorescence microscope (Olympus, Hamburg, Germany) equipped with a 100× oil objective. In all slides analyzed, the numbers of unhybridized cells in the assessed areas were systematically < 1%, and only spots with a similar size, intensity, and shape were counted. The existence of numerical aberrations for the chromosomes studied was considered to be present once the proportion of cells displaying an abnormal number of spots was at percentages higher than the mean value plus three standard deviations of the percentages obtained for that specific chromosome in BM samples from normal controls.7 In addition, the criteria used to define the existence of one or more PC clones was based on the simultaneous use of different combinations of two or three of the four chromosome probes studied.
For all continuous variables included in this study, their mean values, standard deviation, and range were calculated using the SPSS software package (SPSS 8.0 Inc., Chicago, IL); for dichotomous variables, frequencies were reported. Comparisons between groups were performed using either the Mann–Whitney U (binomial, SPSS) or the chi-square tests (SPSS, chi-square) for dichotomous and continuous variables, respectively. Statistical significance was considered to be present when P values < 0.05 were found.
Overall, FISH analysis revealed that 17 of the 30 MGUS cases analyzed (57%) displayed abnormalities for at least 1 of the 4 chromosomes studied, such incidence being slightly lower than that observed in both MM and PCL patients (75% in both cases; P = 0.05). Table 1 summarizes the incidence of numerical abnormalities for each individual chromosome analyzed within each diagnostic group. Among MGUS cases, the most common abnormalities detected were gains of chromosomes 9 and/or 6 (23% and 21% of the patients, respectively) and loss/deletions of chromosomes 13 and/or 17 (21% and 17% of the cases, respectively). Compared with MM patients, MGUS cases were found to have a lower incidence of gains of chromosome 9 (23% vs. 54%, P = 0.002) and monosomy 13/13q- deletions (21% vs. 38%, P = 0.07); with respect to PCL, MGUS patients had a significantly lower incidence of monosomy 13/13q- deletions (21% vs. 75%, P < 0.001) together with a slightly higher frequency of gains of chromosomes 6 (21% vs. 0%, P = 0.05) and 9 (23% vs. 7%, P = 0.1).
|MGUS (n = 30)|
|Diploid||18/24 (75%)||23/30 (77%)||19/24 (79%)||23/30 (77%)|
|Chromosome losses/deletions||1/24 (4%)||0/30 (0%)||5/24 (21%)||5/30 (17%)|
|Chromosome gains||5/24 (21%)||7/30 (23%)||0/24 (0%)||2/30 (7%)|
|MM (n = 158)|
|Diploid||109/142 (77%)||71/156 (45.5%)||89/144 (62%)||136/152 (89%)|
|Chromosome losses/deletions||2/142 (1%)||1/156 (0.5%)||55/144 (38%)||4/152 (3%)|
|Chromosome gains||31/142 (22%)||84/156 (54%)||0/144 (0%)||12/152 (8%)|
|PCL (n = 20)|
|Diploid||13/13 (100%)||14/15 (93%)||5/20 (25%)||13/14 (93%)|
|Chromosome losses/deletions||0/13 (0%)||0/15 (0%)||15/20 (75%)||1/14 (7%)|
|Chromosome gains||0/13 (0%)||1/15 (7%)||0/20 (0%)||0/14 (0%)|
|P value||0.3||< 0.001||< 0.001||0.02|
The simultaneous use of two or more different chromosome probes on purified phenotypically aberrant BM PC allowed us to accurately evaluate the presence of one or more clones of BM PC on the basis of the FISH patterns observed for the four chromosomes analyzed. As shown in Figure 1 and exemplified in Figure 2, two or more different clones of phenotypically aberrant BM PC were detected in a high proportion (67%) of MGUS patients; all remaining cases (33%) displayed a single clone of cells carrying a normal number of copies for the four chromosomes analyzed. The incidence of two or more PC clones was significantly lower among MM (19%) and PCL (15%) cases (Fig. 1). Among cases with ≥ 2 PC clones, upon comparing the MGUS versus MM and PCL subgroups, a marked difference was observed in regard to the incidence of cases in which the abnormal PC clone coexisted with a clone displaying a normal number of copies for the four chromosomes analyzed; this situation was significantly more common in MGUS patients (88%) than in MM (30%) and in PCL (33%) patients (see Table 2 for more detailed information). Moreover, when we focused on cases with a single clone of aberrant PC, once again we observed that in all MGUS patients the PC displayed normal chromosome numbers, whereas in 68% of MM and 65% of PCL patients all the PC carried numerical abnormalities.
Overall, the most frequent abnormal cytogenetic patterns detected in MM for the four chromosomes under analysis consisted of gains of chromosome 9 both in the absence of additional chromosome changes (35 cases, 22%) or with concomitant chromosome losses (31 cases, 20%), as well as monosomy 13/13q- deletions associated with gains of other chromosomes (33 cases, 23%) (Table 3). In contrast, among PCL patients, monosomy 13/13q- deletion alone (13 cases, 65%) represented the most commonly altered cytogenetic pattern (Table 3). It is interesting to note that MGUS patients displayed a rather heterogeneous profile of cytogenetic patterns. Some of them were identical to those found in MM or PCL cases: 13% of MGUS cases demonstrated monosomy 13/13q- deletions alone, 10% had gains of chromosome 9 alone, and 13% displayed gains of chromosome 9 together with gains of chromosome 6. By contrast, there were some specific cytogenetic patterns in MGUS cases that were never detected either in MM or in PCL patients; this situation is represented by the shadow areas in Table 2 for cases with more than one PC clone. Thus, 20% of MGUS patients showed coexistence of two cytogenetically different clones: one carrying monosomy of chromosomes 17 and/or chromosome 6, and the other with a normal number of copies for the four chromosomes analyzed, a pattern never observed in MM patients (Table 3). In turn, PC clones simultaneously carrying monosomy 13/13q- deletions together with gains of chromosomes 9, 6, and/or 17 were exclusively found in MM (23%) and rarely in PCL (5%) but were absent in MGUS.
|Additional chromosome changes||Cytogenetic patterns by FISH|
|Normal diploid (%)||Monosomy 13/13q− (%)||Monosomy 6 or 17 (%)||Gains of chromosome 9 (%)||Gains of chromosomes 6 or 17 (%)|
|MGUS||None||13 (43)||3 (13)||3 (10)||3 (10)||1 (3)|
|Losses||0 (0)||2 (8)||2 (7)||0 (0)||1 (3)|
|Gains||0 (0)||0 (0)||1 (3)||4 (13)||5 (17)|
|MM||None||40 (25)||19 (13)||2 (1)||35 (22)||7 (4.5)|
|Losses||0 (0)||4 (3)||3 (2)||31 (20)||10 (6.5)|
|Gains||0 (0)||33 (23)||2 (1)||25 (16)||25 (16)|
|PCL||None||5 (25)||13 (65)||0 (0)||0 (0)||0 (0)|
|Losses||0 (0)||1 (5)||1 (7)||1 (7)||0 (0)|
|Gains||0 (0)||1 (5)||0 (0)||0 (0)||0 (0)|
|P value||0.13||< 0.001||0.05||< 0.001||0.32|
Conventional karyotypic analysis of clonal PC from MGUS patients has long been hampered by the low numbers at which these predominantly nonproliferating cells are present in BM.25 Because of this, most information currently available on the chromosomal abnormalities present in MGUS derives from the use of interphase FISH.12–14 However, these techniques typically do not provide an overview of all karyotypic abnormalities carried by the clonal PC.10, 26, 27 Thus, information on the genetic alterations of clonal BM PC from MGUS patients remains very limited and restricted to a few chromosomal abnormalities.12–15, 17, 28, 29
Preliminary studies have shown that clonal BM PC from MGUS patients frequently display numerical abnormalities for chromosomes 3, 9, and to a lesser extent for chromosomes 7 and 11,12–14, 20 whereas chromosome 18 is only occasionally altered.12, 18 In the current study, we have shown that once chromosomes 6, 9, 13, and 17 are simultaneously analyzed, numerical chromosomal abnormalities are found in most MGUS cases, but the frequencies of these abnormalities are slightly lower than those observed in MM and PCL. Regarding specific chromosomes, the current study has confirmed13, 14, 20 the relatively high frequency of trisomy 9 (23% of MGUS), and for the first time to our knowledge we have shown that gains of chromosome 6 also are frequent (21%). Despite the finding that monosomy 13/13q- deletions have been associated with an adverse prognosis in multiple myeloma,8, 30–32 several reports have confirmed the presence of this alteration in clonal PC from a variable proportion of MGUS patients,15, 17, 28 as also found in the current study. Interestingly, in the current series, monosomy 17 was found in approximately 17% of the MGUS cases, whereas it was detected at significantly lower frequencies in MM and PCL cases. Although these results were in contrast to those found by the Vienna group,10, 29, 33 we should note that our study was based on immunophenotypically aberrant sorted PC and monosomy 17 was only found in a relatively small subset of these PC.
For many years considerable evidence has existed on the heterogeneous intratumoral cytogenetic nature of plasma cell disorders,34 and many reported cases have illustrated the existence of periods of clonal competition35, 36 before proliferative advantage is reached by one clone.37 In line with these observations, cases also have been reported in which transition of MGUS into active MM was associated with the expansion of a new clone carrying identical idiotypic determinants to those of the original PC clone.38–43 Acquisition of new cytogenetic abnormalities also has been occasionally reported in MM in the transition from stable to progressive disease.44, 45 In turn, Avet-Loiseau et al.15 have suggested that specific patterns of genetic abnormalities could favour development of MGUS into a malignant PC condition. Accordingly, they reported a higher frequency of monosomy 13, but not a different incidence of 14q32 rearrangements, among MM cases who had a prior history of MGUS; a more detailed analysis of both abnormalities showed that the combination of t(4;14)-but not t(11;14)-and monosomy 13/13q- deletions were shared by a significant proportion of MGUS and MM patients.
Based on this background, another goal of our study was to explore the heterogeneity of the PC in MGUS as a way to identify cytogenetic patterns either similar or different from those frequently observed in MM and PCL and that could be of help, in the future, to predict risk of progression of MGUS. For that purpose and to accurately identify small PC clones, highly purified phenotypically aberrant BM PC from MGUS patients were systematically analyzed. This strategy was based on the finding that previous studies have extensively shown that in MGUS, phenotypically aberrant BM PC corresponded to clonal PC, whereas those PC showing a normal phenotype were polyclonal.23 Using this approach, our results showed that pathologic PC from MGUS patients were cytogenetically more heterogenous than clonal PC from MM and PCL, with most MGUS cases showing two or more PC clones. It is interesting to note that among clonal PC from MGUS patients, diploid PC for the four chromosomes analyzed frequently coexisted with cells displaying numerical chromosomal abnormalities, whereas in MM cases with two or more PC clones, usually both of them were aneuploid. Overall, these results supported the notion that among patients with monoclonal gammopathies, aneuploidization occurred in a high proportion of cases at a relatively early stage of the disease, frequently prior to malignancy.
Regarding the specific patterns of aneuploidization that were found, it should be noted that statistically significant differences were observed between MGUS and both MM and PCL. Accordingly, although some FISH patterns were shared by both groups of diseases (i.e., monosomy13/13q- deletions alone, gains of chromosome 9 alone or together with trisomy 6), others were almost exclusively found in either MGUS (i.e., a clone with monosomy 6 and/or 17 together with nuclei from aberrant PC displaying a normal chromosome number) or in MM (i.e., monosomy 13/13q- deletions together with gains of chromosomes 6 and/or 9). To our knowledge, this is the first report in which both unique and similar patterns of aneuploidization were reported in MGUS compared with MM and PCL. Although some of these cytogenetic patterns will potentially never be associated with transformation into MM, others that are found in both disease groups (i.e., isolated trisomy 9 or monosomy 13/13q- deletion) may favour further evolution of MGUS into a malignant disorder; these findings deserve further follow-up prospective studies to confirm this hypothesis.
The results of the current study demonstrated that MGUS patients display a high incidence of numerical chromosome abnormalities that are usually associated with the existence of two or more tumor cell clones. Interestingly, some of the altered interphase FISH patterns found in MGUS patients are extremely rare in MM and PCL and vice versa, whereas others are shared by both disease conditions.
- 10Cytogenetic, interphase, and multicolor fluorescence in situ hybridization analyses in primary plasma cell leukemia: a study of 40 patients at diagnosis, on behalf of the Intergroupe Francophone du Myelome and the Groupe Francais de Cytogenetique Hematologique. Blood. 2001; 97: 822–825., , , et al.
- 21Chronic Leukemia-Myeloma Task Force, National Cancer Institute. Proposed guidelines for protocol studies. II. Plasma cell myeloma. Cancer Chemother Rep. 1973; 4: 145–158.