Clinical significance of cytogenetics and interphase fluorescence in situ hybridization analysis in newly diagnosed multiple myeloma in Taiwan

Original article

Annals of Oncology 16: 1530– 1538, 2005 doi:10.1093/annonc/mdi273 Published online 6 June 2005

Clinical significance of cytogenetics and interphase fluorescence in situ hybridization analysis in newly diagnosed multiple myeloma in Taiwan S.-Y. Huang1, M. Yao1, J.-L. Tang1, W. Tsay1, F.-Y. Lee2, M.-C. Liu2, C.-H. Wang3, Y.-C. Chen3, M.-C. Shen3 & H.-F. Tien1* Departments of 1Internal Medicine, 2Pathology and 3Laboratory Medicine, National Taiwan University Hospital, Taipei, Taiwan Received 3 April 2005; accepted 20 April 2005

Introduction Multiple myeloma (MM) is a genetically unstable malignancy of postgerminal center B-lineage cells [1]. Using conventional cytogenetics (CG) and interphase fluorescence in situ hybridization (FISH), with other modern cytogenetic techniques, it has now been established that aneuploidy with chromosome abnormalities (CA) is almost universal in MM [2 –4]. Almost all CA in MM involve complex numeric and structural changes [2, 5]. Common numeric changes in MM are trisomies involving chromosomes 3, 5, 7, 9, 11, 15, 19 and 21, and monosomies involving chromosomes 8, 13, 16, 17, 22 and Y [2, 5, 6]. Common structural changes in CA include rearrangements of 14q32, partial deletion of chromosome 13,

*Correspondence to: Dr H.-F. Tien, Department of Internal Medicine, National Taiwan University Hospital, No. 7, Chung-Shan South Road, Taipei, Taiwan, ROC. Tel: + 886-2-23970800, ext. 3955; Fax: + 886-2-23959583; E-mail: [email protected] q 2005 European Society for Medical Oncology

duplications of 1q and deletions of 1p, 6q, 11q and 17p [1, 5, 7]. Some CA in MM are not randomly distributed, but are tightly associated and may be responsible for the variability of MM with respect to the natural history, morphological and immunological features of myeloma cells and clinical status [5, 6, 8, 9]. Two aneuploidy groups of MM, hyperdiploid and non-hyperdiploid MM, have been proposed based on chromosome numbers [10, 11]. These two ploidy subcategories are associated with distinct clinical presentation and outcome [10–12]. Analogous to acute and chronic leukemias, such CA of MM represent not only the biological basis of the clinical heterogeneity, but also the prognostic parameters [13]. Hypodiploidy and deletion of chromosome 13, including monosomy 13 and partial deletion of 13q (D13), could be used as negative prognostic predictors for MM patients receiving either conventional chemotherapy [12, 14–16] or high-dose chemotherapy followed by autologous hematopoietic stem cell transplantation (auto-HSCT) [17–21]. Other recurrent CA in

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

Background: The incidence of multiple myeloma (MM) is lower in Asia than in Western countries. However, it is not known whether cytogenetic abnormalities (CA) characteristic of MM in Asia differ from those documented in the West. Patients and methods: We analyzed CA by conventional cytogenetics (CG) and/or fluorescence in situ hybridization (FISH), assessed their clinical significance in 150 Chinese MM patients and compared our data with that derived from Western countries. Results: CA were detected by CG (CG_CA) in 44 (29.3%) of the 150 patients and by FISH (FISH_CA) in 59 (67%) of the 88 patients studied. Presence of either CG_CA or FISH_CA was associated with a poor prognosis. Patients with CG_CA and hyperdiploid chromosomes, always associated with several trisomies, had a longer survival (median 25 months versus 12 months; P = 0.025) in comparison with those with non-hyperdiploid chromosomes, usually associated with a monosomy 13/partial deletion of 13q (D13) and a rearrangement of 14q32. A novel recurrent CG_CA, add(19)(p13), was found in four patients: all males with immunoglobulin G/l isotypes, extramedullary myeloma at diagnosis and a poor prognosis. Three groups of patients with significantly different survival, CG_D13, FISH_D13 but without CG_D13, and neither CG_D13 nor FISH_D13 (median 9 versus 15 versus 32 months; P = 0.013) were identified. Conclusions: We conclude that MM CA in our patients are similar to those noted in Western countries, and that combined CG and FISH analysis can predict prognosis. The clinical significance of add(19)(p13) needs to be further investigated. Key words: cytogenetics, fluorescence in situ hybridization, multiple myeloma, prognosis

1531

Patients and methods

anti-human kappa and anti-human lambda probes, conjugated with 7amino-4-methylcourmarin-3-acetic acid; Vector Laboratories, Burlingame, CA, USA) as described previously [32]. BM cells from five transplantation donors were used as normal controls. Threshold levels for gain or loss of signals for each probe in cIg-FISH, which were set at the mean of normal controls plus three standard deviations, were as follows: 13q14 deletion, 1.8%; 17p13 deletion, 2.4%; + 3, 0.7%; + 7, 2.0%; + 11, 1.9%; and + 18, 1.9%. However, we only regarded the findings of any abnormal signals in more than 10% of 100–300 scored PC as true evidence for CA on the cIg-FISH analysis, in order to avoid false positives [11].

Treatment A total of 131 (87.3%) received chemotherapy with MP (melphalan 9 mg/m2 and prednisolone 60 mg/m2 daily orally on days 1–4) regimen every 4 –6 weeks until plateau phase was achieved or the disease was refractory to the treatment. Owing to poor clinical status or hesitation, 19 patients (12.7%) did not receive chemotherapy. A total of 71 patients (47.3%) with high tumor burden were treated with two to four cycles of combination chemotherapy of VAD [vincristine 0.4 mg/m2, continuous intravenous infusion (CIVF), days 1– 4; doxorubicin, 9 mg/m2, CIVF, days 1–4; dexamethasone, 40 mg, CIVF, days 1–4 and 8– 11] at an interval of 4 weeks. High-dose chemotherapy (melphalan 200 mg/m2) followed by auto-HSCT was administered to 10 patients.

Patients Between June 1986 and December 2003, 150 newly diagnosed and untreated MM patients with accessible cytogenetic results at our hospital were enrolled. In addition to the CG analysis, 88 patients had FISH performed simultaneously. The work was approved by our institutional human ethics committee that oversees research involving human subjects. Written informed consent was obtained from all patients.

Treatment response Treatment response criteria were as described previously [33, 34]. In brief, responders included patients who had achieved a complete response, partial response or minimal response. Non-responders included the patients who had no response or had progressive disease. Overall survival (OS) was defined as the time period from the date of diagnosis to the date of death, regardless of cause.

Conventional cytogenetics Bone marrow (BM) samples were aspirated into heparinized syringes and CG with G-banding method was performed on BM cells after 1– 3 days of unstimulated culture as described previously [26]. CA and ploidy levels were defined according to International System for Human Cytogenetic Nomenclature [30]. For comparison with the results in other areas [10], patients were divided into two groups: one with hyperdiploid MM (47–57 chromosomes) and the other with non-hyperdiploid MM including the hypodiploid (35–45 chromosomes), pseudodiploid (46 chromosomes) and tri-/tetraploid MM (58–103 chromosomes). Complex CA was defined by more than two cytogenetic changes.

Fluorescence in situ hybridization FISH analysis was performed as previously described [31]. Deletion of 13q14 was determined by combined use of a retinoblastoma gene-1 probe LSIRB1 (Vysis, Downers Grove, IL, USA) and a reference chromosome 10 centromeric probe LPE010G (Cytocell Ltd, Banbury, UK) in dual-color. Deletion of 17p13, where the p53 gene was located, was determined by use of a LSIp53 probe (Vysis) combined with a chromosome 17 centromeric probe (Cytocell Ltd). Several other centromeric probes, LPE003R, LPE007G, LPE011R and LPE018G (Cytocell Ltd), were used to detect trisomies/monosomies of chromosomes 3, 7, 11 and 18, respectively. To improve the specificity of the FISH analysis, we combined the FISH technique with immunofluorescence staining of the cytoplasmic light chain of plasma cell (PC) (cIg-FISH;

Statistical analysis x2 or Fisher’s exact tests were used for between-group comparison of the discrete variables. Two-sample t-test was used for between-group comparison of the means. Kaplan–Meier survival curves were used for estimation of OS. Log-rank test was used to test for differences in OS between groups. Several salient clinical and laboratory variables, including age, sex, disease stage, BM plasmacytosis, M-component isotype, levels of hemoglobin (Hb), white blood cell (WBC), platelet (PLA), lactate dehydrogenase (LDH), calcium (Ca), creatinine (Cr), C-reactive protein (CRP) and b2-microglobulin (b2M) were assessed in all patients at diagnosis to determine its possible association with CA. All the variables and the CA were examined for their prognostic values on OS. Those factors with statistically prognostic significance from univariate analysis were tested by multivariate analysis with the Cox proportional hazards regression model using forward stepwise selection. In these prognostic analyses, continuous variables were categorized by the cut_ 60 years, Durie–Salmon stage > _ III, BM off values as follows: age > _ 30%, IgA isotype, Hb <10 g/dl, WBC <4  109/l, PLA plasmacytosis > _ 465 IU/l, Ca > _ 2.4 mmol/l, Cr > _ 2 mg/dl, CRP <1.5  1011/l, LDH > > _ 4 mg/dl and b2M > _ 2.5 mg/l, as set up in previous reports [18, 20]. All _ 0.05 condirectional P values were two-tailed, with a P value of < sidered significant for all tests. All analyses were performed using SPSS 8.0 software (SPSS, Inc., Chicago, IL, USA).

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

MM, such as t(4;14), t(14;16), deletion of 17p13, abnormalities of 11q and 22q, monosomies of 2, 3, 14 and 19, and 1p deletions, are associated with a poor prognosis [6, 14, 16, 22– 24]. Analysis of CA in MM patients is required to elucidate the pathogenesis of the disease, predict treatment outcome and possibly identify distinct groups of MM patients who could benefit from aggressive or novel therapies. Uneven geographical distribution of non-random CA in malignant disorders has been reported previously [25, 26]. It was suggested that the heterogeneity in the incidence of nonrandom CA in various areas was a reflection of ethnic differences or environmental factors [27]. We found that CA in Chinese lymphoma patients, in Taiwan, were different from those noted in lymphoma patients in Western countries [28]. Indeed, the incidence of MM in Asia is much lower than that recorded in Western countries [29]. It is not clear whether this disparity in CA between Asia and Western countries also extends to CA characteristic of MM. Therefore, we have used conventional cytogenetics and FISH to analyze CA and their clinical significance in 150 Chinese patients with newly diagnosed MM in Taiwan and compared our findings with those derived from Western countries.

1532

Results Patient characteristics There were 98 males and 52 females, with a median age of 62 years (range 24– 88). The main bioclinical features are summarized in Table 1. Up to December 2004, the median follow-up time for the patients was 62 months. The median OS for all patients was 27 months.

Prevalence and details of CA detected by CG

Table 1. Patient characteristics and demographics Characteristics

All patients (n = 150)

With CA (n = 44)

Without CA (n = 106)

Sex (male/female)

98/52

28/16

70/36

0.851

Age (years)a

62.0 ± 13.9

63.3 ± 14.4

61.3 ± 13.7

0.415

I and II

41 (27.3)

10 (22.7)

31 (29.2)

IIIA and IIIB

109 (72.7)

34 (77.3)

75 (70.8)

72 (48.0)

21 (47.7)

51 (48.1)

Durie–Salmon stage (%)

P value (two-tailed)

0.546

Isotype (%)

0.772

IgG IgA

38 (25.3)

13 (29.5)

25 (23.6)

Light-chain

29 (19.3)

6 (13.6)

23 (21.7)

Othersb

11 (7.4)

4 (9.0)

7 (6.6)

Hb (gm/dl)a

8.8 ± 2.6

8.2 ± 2.3

9.1 ± 2.7

0.063

a

WBC (10 /l)

6.9 ± 5.0

8.1 ± 8.0

6.4 ± 2.9

0.176

PLA (1011/l)a

1.8 ± 0.9

1.5 ± 0.8

1.9 ± 1.0

0.010

Cr (mg/dl)

2.7 ± 2.6

2.5 ± 2.1

2.7 ± 2.7

0.677

Ca (mmol/l)a

2.4 ± 0.5

2.5 ± 0.5

2.4 ± 0.4

0.164

LDH (IU/l)

519 ± 460

657 ± 698

461 ± 297

0.081

CRP (mg/dl)a

2.0 ± 3.3

1.8 ± 3.2

2.1 ± 3.3

0.653

b2M (mg/l)

8.5 ± 11.5

7.4 ± 7.8

9.0 ± 12.7

Plasma cell (BM, %)a

48.8 ± 28.2

64.3 ± 28.1

42.4 ± 25.8

Extra-MM (%)c

30 (20.0)

15 (34.1)

15 (14.2)

0.008

Chemotherapyd

131

39

92

1.000

9

a

a

a

0.512 < 0.001

VAD

71

26

45

0.074

Responders (%)

75 (57.3)

20 (51.3)

55 (59.8)

0.441

Non-responders (%)

56 (42.7)

19 (48.7)

37 (40.2)

Auto-HSCT

10

4

6

a

0.480

Mean ± SD. Including MM with IgD, IgM or non-secretory isotypes. c Manifestation of extramedullary myeloma at diagnosis. d Number of patients who had received conventional chemotherapy; responders included patients who had minimal response, partial response or complete response ever; non-responders included patients who had no change or progressive disease (see text). CA, cytogenetic abnormalities; Hb, hemoglobin; WBC, white blood cell count; PLA, platelet; Cr, creatinine; Ca, calcium; LDH, lactate dehydrogenase; CRP, C-reactive protein; b2M, b2-microglobulin; BM, bone marrow; Extra-MM, extramedullary myeloma; VAD, vincristine, doxorubicin, dexamethasone; Auto-HSCT, autologous hematopoietic stem cell transplantation. b

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

Among the 150 newly diagnosed MM patients, CA were detected by CG (CG_CA) in 44 patients (29.3%; Table 2). The prevalence of CA increased with progressing disease stage, being 14.3%, 29.6% and 31.2% in the patients with Durie–Salmon stage I, II and III diseases, respectively. Of the

44 patients with CG_CA, 19 (43.2%) showed hyperdiploidy and 25 (56.8%) non-hyperdiploidy; the latter included 11 (25%) hypodiploidy, 11 (25%) pseudodiploidy and three (6.8%) tri-/tetraploidy (Table 2). A total of five patients had numeric abnormalities only, nine had structural abnormalities only and the remaining 30 patients exhibited both numeric and structural abnormalities. Complex CA were found in 35 patients (79.5%) with a median of 12 changes per patient (range four to 24). Common numerical changes were trisomies 9 ( + 9, 18/44 or 40.9%), 15 (34.1%), 3 (25%), 7 (25%) and 19 (25%), and the monosomy 13 (–13; 31.8%). At least one trisomy was found in 24 patients (24/44; 54.5%); trisomies tended to occur in clusters, and 18 of these 24 patients had multiple trisomies (more than three trisomies) (Table 2). The incidence of multiple trisomies was higher in hyperdiploid than non-hyperdiploid MM (84.2%

1533 Table 2. Cytogenetic data from 150 multiple myeloma patients Karyotype

Number of patients

Incidence in patients with CA (%)

Incidence in MM patients (%)

Abnormal

44

100

29.3

Hyperdiploidy

19

43.2

12.7

Hypodiploidy

11

25.0

7.4

Pseudodiploidy

11

25.0

7.4

Tri-/tetraploidy

6.8

2.0

18a

40.9

12.0

–13/13q14 –

19b

43.2

12.7

1q duplication

16

36.4

10.7

14q32

13

29.5

8.7

1p –

12

27.3

8.0

11q13 or 23

12

27.3

8.0

–17/17p –

8c

18.2

5.3

8q24

7

15.9

4.7

7

15.9

4.7

6q

–

a

With a median of seven trisomies (range three to 11); 13 patients had four or more trisomies. b Including 14 patients with 13 and five with deletion of long arm of chromosome 13 involving band q14; 10 of them had simultaneously 1q duplication, seven each had 14q32 rearrangements and 1p – and six each had 11q13 or 23 rearrangements and 17/17p – . c Including four patients with 17 and four with 17p – . CA, chromosome abnormality; MM, multiple myeloma.

versus 8.0%, P <0.001). The most frequent structural aberrations were duplications of 1q (16/44, 36.4%), followed by rearrangements involving 14q32 (29.5%), deletions of 1p (1p – , 27.3%), rearrangements involving 11q13 (15.9%) or 11q23 (11.4%), abnormalities involving 8q24 (15.9%) and interstitial deletions of 6q (15.9%). CA involving chromosome regions where immunoglobulin light-chain genes were located, 2p11 (k) and 22q11 (l), were noted in only one patient (2.3%) each. CA of chromosome 13 were observed in 23 patients, including 13 in 14 patients and partial deletion involving 13q14 in five patients; thus, a total of 19 patients showed loss of chromosome 13q14 (Table 2). With regards to abnormalities usually accompanied with other structural aberrations, 10 of the 19 patients had simultaneous 1q duplication, seven each had 14q32 rearrangements and 1p – , and six each had 11q rearrangements and 17/17p13 – . Interestingly, the incidence of 13 or partial deletion of 13q (D13) detected by the CG (CG_D13) was more common in the non-hyperdiploid MM than in the hyperdiploid MM (56% versus 26.3%; P = 0.049). 14q32 rearrangements detected by the CG (CG_14q32) were observed in 13 patients (13/44; 29.5%), including four with t(11;14)(q13;q32), one with t(6;14)(p21;q32) and eight with add(14)(q32), in which the translocation partners of 14q32 could not be identified. Similar to the CG_D13, the incidence of CG_14q32 was also

FISH is more sensitive than CG in detecting specific CA Data from the comparisons between CG and FISH in the detection of specific CA in the 88 patients who had both techniques simultaneously performed is shown in Table 3. CG_CA were noted in 23 patients (26.1%) and CA detected by FISH (FISH_CA) were noted in 59 patients (67%). FISH_CA from three representative patients are shown in Figure 1. Interestingly, 40 of the 65 patients who did not have CG_CA showed FISH_CA. Deletion of 13q14 on FISH analysis (FISH_D13) occurred in 30 patients. Notably, the FISH_D13 was all monoallelic and present in all but one patient with CG_D13.

Bioclinical features of patients with the CA The salient characteristics in the patients with CG_CA and those with normal karyotypes were compared and the results are presented in Table 1. Significantly, the patients with CG_CA had more plasma cells in the BM, lower PLA levels and higher incidence of extramedullary myeloma (extra-MM) at diagnosis. Among the 44 patients with CG_CA, the patients with CG_D13 had lower Hb and PLA levels in comparison to other patients in this group (median 7.2 versus 9 g/dl, P = 0.010; and 1.2  1011/l versus 1.7  1011/l, P = 0.048, respectively). The incidence of this abnormality was higher in patients with a l subtype MM than in those with a k subtype (83.3% versus 12%; P <0.001). Patients with CA involving 8q also had lower platelet levels but higher calcium levels in comparison with patients without this CA (median 0.9  1011/l versus 1.6  1011/l, P = 0.007; and 2.8 versus 2.4 mmol/l, P = 0.035, respectively). Patients with CA involving 1p were more likely to have soft tissue plasmacytomas in comparison with patients without this CA (38.9% versus 7.7%; P = 0.021). Interestingly, all four patients with 19p13 rearrangements, add(19)(p13), Table 3. Comparison of CG and FISH data Chromosome abnormality

Total

No. of patients with the abnormality (%) CG

FISH

23 (26.1)

59 (67.0)

a

8 (9.1)

30 (34.1)

del(17p13)b

3 (3.4)

6 (6.8)

+3

8 (9.1)

23 (26.1)

+7

6 (6.8)

19 (21.6)

+ 11

5 (5.7)

18 (20.5)

+ 18

1 (1.1)

14 (15.9)

del(13q14)

a

Including 13/13q14 – . Including 17/17p13 – . CG, conventional cytogenetics; FISH, fluorescence in situ hybridization. b

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

3

_ 3) Multiple trisomies (>

higher in the non-hyperdiploid MM than in the hyperdiploid MM (40% versus 15.8%; P = 0.081).

1534

were males. The add(19)(p13) was the sole chromosome abnormality in one patient. These patients all had monoclonal immunoglobulin of IgG/l isotype and showed extra-MM at diagnosis, compared with 15% and 27.5%, respectively, in those patients with non-add(19)(p13) CA in this study (P = 0.037 and 0.01, respectively). Of the 88 patients who underwent FISH analysis, the differences of the bioclinical characteristics between the patients with the specific FISH_CA and those without were also compared. Patients with FISH_D13 had a higher incidence of extra-MM (P = 0.02) at diagnosis. Additionally, the incidence of trisomy 11 in patients was associated with IgG MM (P = 0.042).

Association between CG_CA and clinical outcome Notably, the patients with CG_CA had a significantly shorter median OS than the patients with normal karyotypes (18 ± 4.2 versus 28 ± 5.1 months; P = 0.029) (Figure 2A). In univariate analysis, other clinical and laboratory variables that were

Association between FISH_CA and clinical outcome Among the 88 patients who had undertaken the CG and FISH analysis simultaneously, patients with FISH_D13 had a shorter OS in comparison with patients without this CA (13 ± 2.2 versus 32 ± 9 months; P = 0.014) (Figure 2E). FISH_D13, BM _ 30%, LDH > _ 465 IU/l, Cr > _ 2 mg/dl, CRP plasmacytosis > > _ 4 mg/dl and response to chemotherapy chosen from univariate analysis were tested in subsequent multivariate analysis. The CA of FISH_D13 was the best independent predictor of a shorter OS (RR of death 3.7; 95% CI 2 –7; P <0.001). Another prognostic factor that remained significant in the multivariate _ 2 mg/dl (RR of death 2.6; 95% CI 1.4–4.9; analysis was Cr > P = 0.004). Interestingly, significant differences in OS could also be demonstrated among the three groups of patients with

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

Figure 1. Combined fluorescence in situ hybridization (FISH) and immunofluorescence staining of the cytoplasmic light chain of interphase cells. The plasma cell was distinguishable by the intense blue fluorescence of the cytoplasm. (A) A plasma cell from a normal control showing two green signals for the centromeres of chromosome 10 (LPE010G) and two red signals for 13q14 (LSIRB1). The surrounding non-plasma cells do not have blue cytoplasmic staining. (B) A plasma cell from a multiple myeloma (MM) patient with monosomy 13/partial deletion of chromosome 13q showing two green signals for the centromeres of chromosome 10 but only one red signal for 13q14. (C) A plasma cell from a MM patient without abnormality of chromosomes 11 and 18 showing two red signals for the centromeres of chromosome 11 (LPE011R) and two green signals for the centromeres of chromosome 18 (LPE018G). (D) A plasma cell from a MM patient with trisomy 11 showing two green signals for chromosome 18 but three red signals for chromosome 11 (600 magnification; fluorescence microscope from Axioplan, Zeiss, Germany).

associated with shorter OS were stage III disease (P = 0.009), _ 30% (P = 0.006), Hb <10 g/dl (P = 0.006), BM plasmacytosis > _ 465 IU/l (P = 0.032), Cr PLA <1.5  1011/l (P <0.001), LDH > > _ 2 mg/dl (P <0.001), CRP > _ 4 mg/dl (P = 0.006) and b2M > _ 2.5 mg/l (P = 0.012). In multivariate analysis, only the CG_CA [relative risk (RR) of death 1.9; 95% confidence interval (CI) 1.2–3.2; P = 0.009], PLA <1.5  1011/l (RR of _ 2 mg/dl (RR death 2.1; 95% CI 1.3– 3.4; P = 0.001) and Cr > of death 2.5; 95% CI 1.5– 4; P <0.001) were the independent prognostic factors for OS. CG_CA remained as an independent prognostic factor on OS only if the patients who had received chemotherapy were analyzed and the response to chemotherapy, as a bimodal covariate (responders versus nonresponders), was also included in the multivariate analysis (data not shown). Among the 44 patients with the CG_CA, the median OS for patients with hyperdiploidy, pseudodiploidy, hypodiploidy and tri-/tetraploidy was 25 ± 12.8, 18 ± 8.1, 9 ± 5.2 and 7 ± 0 months, respectively (P = 0.05). Patients with hyperdiploid MM had a significantly longer median survival in comparison with non-hyperdiploid MM patients (25 ± 12.8 versus 12 ± 4 months; P = 0.025) (Figure 2B). Furthermore, among the patients with CG_CA, patients with CG_D13 had a significantly shorter median survival in comparison with patients without this CA (9 ± 3.5 versus 22 ± 13.5 months; P = 0.047) (Figure 2C). In addition, the four patients with add(19)(p13) were found to have much shorter OS than the patients with other CA (2.5 ± 2.5 versus 18 ± 5.8 months; P = 0.032) (Figure 2D). No other recurrent CG_CA showed a significant impact on OS among patients with CG_CA. To determine the most important factor contributing to shortened survival in patients with CG_CA, the ploidy status (hyperdiploidy versus non-hyperdiploidy), CG_D13, add(19)(p13), as well as IgA _ 2 mg/dl that were associisotype, PLA <1.5  1011/l and Cr > ated with shorter OS by univariate analysis, were subsequently tested by multivariate analysis. Our data showed that only _ 2 mg/dl and PLA <1.5  1011/l were indeadd(19)(p13), Cr > pendent prognostic factors, and the associated RRs of death were 7.2 (95% CI 2– 25.8), 3.4 (95% CI 1.3–8.7) and 2.9 (95% CI 1.3–6.5), respectively (P = 0.002, 0.011 and 0.012, respectively).

1535

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

Figure 2. Kaplan–Meier survival curves of the multiple myeloma (MM) patients showing (A) the overall survival in the patients with normal karyotypes is longer than those with chromosomal abnormalities (CA) detected by conventional cytogenetics (CG_CA); (B) among the 44 patients with CG_CA, the overall survival in the hyperdiploid group is longer than in the non-hyperdiploid group; (C) patients with CA other than CA_D13 had better prognosis in comparison with patients with CA_D13; (D) the patients with CA other than add(19)(p13) had better overall survival in comparison to patients with add(19)(p13); (E) in the subgroup analysis of 88 patients who had undertaken CG and FISH analyses simultaneously, the overall survival of patients without FISH_D13 was longer than that of patients with FISH_D13; (F) significant differences in survival were noted between patients with CG_D13, patients with FISH_D13 only but without CG_D13, and other patients without CG_D13 or FISH_D13.

1536 CG_D13, FISH_D13 only without CG_D13, and other karyotypes with neither CG_D13 nor FISH_D13 (9 ± 7.8, 15 ± 2.8 and 32 ± 8.9 months, respectively; P = 0.013) (Figure 2F).

Discussion

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

The prevalence of CG_CA was 29.3% in the cohort of 150 Chinese patients with newly diagnosed MM. The majority of CG_CA cases showed complex structural and numerical changes, and involved hyperdiploid chromosomes. In general, the incidence of CG_CA and the frequent non-random abnormalities in our study were comparable to those from Western countries [2, 5, 6, 10 –12, 20, 22]. In line with the findings that patients with CG_CA had a larger tumor burden, a higher labeling index and Ki-67 growth fraction of PC when compared with the patients without CG_CA [6, 7, 35, 36], we found that the prevalence of CG_CA was higher in stage III than in stage I disease, and the patients with CG_CA had significantly higher BM plasmacytosis, lower levels of platelets and higher incidence of extra-MM than the patients with normal karyotypes. We demonstrated that the CG_CA in MM was associated with a shorter OS. Although CG_CA was associated with a high proliferative index of PC and advanced tumor stage, it was still an independent prognostic factor on multivariate analysis. In addition, there were distinct differences in clinical and laboratory features between hyperdiploid and non-hyperdiploid MM. The former was more frequently associated with CG_CA of multiple trisomies, whereas the latter was more associated with the CG_D13 and CG_14q32. Importantly, the median OS was significantly shorter in the patients with nonhyperdiploid MM than in those with hyperdiploid MM (12 versus 25 months), especially in those with hypodiploid (9 months) and tri-/tetraploid chromosomes (7 months), which were considered as the same diploidy status [12]. Neither CG_D13 nor CG_14q32 could be used to further separate non-hyperdiploidy MM into different prognostic categories. These findings are similar to reports from Western countries [10–12], and suggest that the two distinct groups of MM, delineated by the chromosome ploidy and patterns of CG_CA, truly exist without ethnic or geographical variations. However, the pathogenesis and mechanism for the survival differences between the two groups are still obscure. Unfortunately, further studies are partly hampered by lack of suitable cell lines derived from the hyperdiploid MM [11]. We recently developed a new SCID mouse/human chimeric animal model, which can sustain the viability and growth of primary MM cells and might be beneficial in facilitating further studies [37]. Owing to the low proliferative index of PC, CG used alone underestimates the prevalence of CA in MM [2, 11]. FISH enables identification of CA without requiring metaphases within the PC. However, FISH can only provide information for specific target regions, which is of limited value in the interpretation of highly complex karyotypes in MM. In contrast, CG gives a complete overview of CA present in the

malignant clone. Therefore, combining the two methods is particularly useful in the analysis of clinically relevant CA in MM. In this study, patients with FISH_D13 had a shorter median survival than those without this CA, and FISH_D13 remained the best independent prognostic factor on the multivariate analysis. Furthermore, there was a significant difference in the OS among patients with CG_D13, patients with FISH_D13 but without CG_D13, and patients with neither CG_D13 nor FISH_D13 (median 9 versus 15 versus 32 months, respectively; P = 0.013). It is likely that the net effect of D13 on prognosis is greater when D13 is detected by CG than when it is detected by FISH [21, 36]. This is because of the additive effects of a large tumor burden and a high proliferation rate of tumor cells, which usually accompany CG_CA, on the prognosis [11]. To the best of our knowledge, this is the first report to show that MM with add(19)(p13) is a distinct cytogenetic entity that is associated with the male gender, IgG/l isotype, presence of extra-MM and a poor prognosis. To further support the notion, two other previously treated MM patients, not included in this study, who had complex chromosome changes with add(19)(p13) at the time of referral from other hospitals were also male, had IgG/l isotype and showed extra-MM at that time (personal communication). From a review of available literature, non-random chromosomal rearrangements of 19p13 have been reported in four of 21 MM patients (19%) from Japan [38]. One of these patients had add(19)(p13) and the other three had translocations involving 19p13. Additionally, all four patients had comparable clinical features to our patients, all were males with extra-MM and two of them had IgG/l isotypes [38]. Notably, among the patients with CG_CA, the patients with add(19)(p13) were likely to have much shorter OS than the patients without this CA. Rather than the CG_D13 and nonhyperdiploidy, the presence of add(19)(p13) became the most powerful prognostic factor associated with short OS on multivariate analysis in the patients with CG_CA. However, the prognostic impact of add(19)(p13) in MM needs to be confirmed by further studies involving a larger number of patients. The genetic changes associated with the chromosomal abnormalities remain to be elucidated. Several known genes mapped to the band 19p13.3, such as the E2A gene, which encodes the enhancer-binding protein E12/E47 and is rearranged in most cases of acute lymphoblastic leukemia with translocation t(1;19)(q23;p13.3) [39], and the basigin gene, which is a member of the Ig superfamily and plays a role in intercellular recognition [40], are potential candidates. Recently, a novel cryptic translocation involving 19p13.3 and IgH was found in chronic B-cell lymphocytic leukemia and large B-cell lymphoma [41], which hinted at the involvement of other novel genes. In summary, CG and FISH analyses identified several chromosomal abnormalities highly associated with the clinical features and prognosis of newly diagnosed MM patients. Combined CG and FISH studies separated MM patients into three groups with significantly different OS; one with CG_D13,

1537 one with FISH_D13 but not CG_D13, and the remaining one with neither CG_D13 nor FISH_D13. The presence of add(19)(p13) as the sole chromosomal abnormality in one patient suggested that this CA plays a role in pathogenesis for a subset of MM patients and demonstrated the existence of a specific cytogenetic entity.

15.

16.

Acknowledgements This study was supported in part by grants NSC90-2314B-002-268 and NSC91-2314-B-002-131 from the National Science Council of the Republic of China.

17.

18.

1. Bergsagel PL, Kuehl WM. Chromosome translocations in multiple myeloma. Oncogene 2001; 20: 5611–5622. 2. Zandecki M, Lai JL, Facon T. Multiple myeloma: almost all patients are cytogenetically abnormal. Br J Haematol 1996; 94: 217 –227. 3. Cigudosa JC, Rao PH, Calasanz MJ et al. Characterization of nonrandom chromosomal gains and losses in multiple myeloma by comparative genomic hybridization. Blood 1998; 91: 3007–3010. 4. Sawyer JR, Lukacs JL, Thomas EL et al. Multicolour spectral karyotyping identifies new translocations and a recurring pathway for chromosome loss in multiple myeloma. Br J Haematol 2001; 112: 167 –174. 5. Nilsson T, Ho¨glund M, Lenhoff S et al. A pooled analysis of karyotypic patterns, breakpoints and imbalances in 783 cytogenetically abnormal multiple myelomas reveals frequently involved chromosome segments as well as significant age- and sex-related differences. Br J Haematol 2003; 120: 960–969. 6. Debes-Marun CS, Dewald GW, Bryant S et al. Chromosome abnormalities clustering and its implications for pathogenesis and prognosis in myeloma. Leukemia 2003; 17: 427–436. 7. Calasanz MJ, Cigudosa JC, Odero MD et al. Cytogenetic analysis of 280 patients with multiple myeloma and related disorders: primary breakpoints and clinical correlations. Genes Chromosom Cancer 1997; 18: 84– 93. 8. Avet-Loiseau H, Facon T, Grosbois B et al. Oncogenesis of multiple myeloma: 14q32 and 13q chromosomal abnormalities are not randomly distributed, but correlate with natural history, immunological features, and clinical presentation. Blood 2002; 99: 2185–2191. 9. Garand R, Avet-Loiseau H, Accard F et al. t(11;14) and t(4;14) translocations correlated with mature lymphoplasmacytoid and immature morphology, respectively, in multiple myeloma. Leukemia 2003; 17: 2032–2035. 10. Smadja NV, Fruchart C, Isnard F et al. Chromosomal analysis in multiple myeloma: cytogenetic evidence of two different diseases. Leukemia 1998; 12: 960–969. 11. Fonseca R, Barlogie B, Bataille R et al. Genetics and cytogenetics of multiple myeloma: A workshop report. Cancer Res 2004; 64: 1546–1558. 12. Smadja NV, Bastard C, Brigaudeau C et al. Hypodiploidy is a major prognostic factor in multiple myeloma. Blood 2001; 98: 2229–2238. 13. Barlogie B, Shaughnessy J, Tricot G et al. Treatment of multiple myeloma. Blood 2004; 103: 20–32. 14. Calasanz MJ, Cigudosa JC, Odero MD et al. Hypodiploidy and 22q11 rearrangements at diagnosis are associated with poor

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29. 30. 31.

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

References

prognosis in patients with multiple myeloma. Br J Haematol 1997; 98: 418–425. Fonseca R, Harrington D, Oken MM et al. Biological and prognostic significance of interphase fluorescence in situ hybridization detection of chromosome 13 abnormalities (D13) in multiple myeloma: An Eastern Cooperative Oncology Group study. Cancer Res 2002; 62: 715–720. Ko¨nigsberg R, Zojer N, Ackermann J et al. Predictive role of interphase cytogenetics for survival of patients with multiple myeloma. J Clin Oncol 2000; 18: 804–812. Desikan R, Barlogie B, Sawyer J et al. Results of high-dose therapy for 1000 patients with multiple myeloma: Durable complete remissions and superior survival in the absence of chromosome 13 abnormalities. Blood 2000; 95: 4008–4010. Fancon T, Avet-Loiseau H, Guillerm G et al. Chromosome 13 abnormalities identified by FISH analysis and serum b2-microglobulin produce a powerful myeloma staging system for patients receiving high-dose therapy. Blood 2001; 97: 1566–1571. Fassas AB, Spencer T, Sawyer J et al. Both hypodiploidy and deletion of chromosome 13 independently confer poor prognosis in multiple myeloma. Br J Haematol 2002; 118: 1041–1047. Shaughnessy J, Jacobson J, Sawyer J et al. Continuous absence of metaphase-defined cytogenetic abnormalities, especially of chromosome 13 and hypodiploidy, ensures long-term survival in multiple myeloma treated with Total Therapy I: interpretation in the context of global gene expression. Blood 2003; 101: 3849–3856. Shaughnessy J Jr, Tian E, Sawyer J et al. Prognostic impact of cytogenetic and interphase fluorescence in situ hybridization-defined chromosome 13 deletion in multiple myeloma: early results of total therapy II. Br J Haematol 2003; 120: 44–52. Tricot G, Barlogie B, Jagannath S et al. Poor prognosis in multiple myeloma is associated only with partial or complete deletions of chromosome 13 or abnormalities involving 11q and not with other karyotype abnormalities. Blood 1995; 86: 4250–4256. Drach J, Ackermann J, Fritz E et al. Presence of a p53 gene deletion in patients with multiple myeloma predicts for short survival after conventional-dose chemotherapy. Blood 1998; 92: 802–809. Chang H, Sloan S, Li D et al. The t(4;14) is associated with poor prognosis in myeloma patients undergoing autologous stem cell transplant. Br J Haematol 2004; 125: 64– 68. Johansson B, Mertens F, Mitelman F. Geographic heterogeneity of neoplasia-associated chromosome aberrations. Genes Chromsome Cancer 1991; 3: 1 –7. Tien HF, Wang CH, Lin MT et al. Correlation of cytogenetic results with immunophenotype, genotype, clinical features and ras mutation in acute myeloid leukemia: a study of 235 Chinese patients in Taiwan. Cancer Genet Cytogenet 1995; 84: 60– 68. Mitelmen F. Geographic heterogeneity of chromosome aberrations in hematologic disorders. Cancer Genet Cytogenet 1986; 20: 203–208. Chen CY, Yao M, Tang JL et al. Chromosomal abnormalities of 200 Chinese patients with non-Hodgkin’s lymphoma in Taiwan: with special reference to T-cell lymphoma. Ann Oncol 2004; 15: 1091– 1096. Riedel DA, Pottern LM. The epidemiology of multiple myeloma. Hematol Oncol Clin North Am 1992; 6: 225– 247. Mitelman F (ed.). ISCN. An International System for Human Cytogenetic Nomenclature. Basel, Switzerland: S Karger 1995. Ko BS, Tang JL, Lee FY et al. Additional chromosomal abnormalities and variability of BCR breakpoints in Philadelphia chromosome/BCR-ABL-positive acute lymphoblastic leukemia in Taiwan. Am J Hematol 2002; 71: 291 –299.

1538 37. Huang SY, Tien HF, Su FS, Hsu SM. Non-irradiated NOD/SCIDhuman chimeric animal model for primary human multiple myeloma—a potential in vivo culture system. Am J Pathol 2004; 164: 747–756. 38. Taniwaki M, Nishida K, Takashima T et al. Nonrandom chromosomal rearrangements of 14q32.3 and 19p13.3 and preferential deletion of 1p in 21 patients with multiple myeloma and plasma cell leukemia. Blood 1994; 84: 2283– 2290. 39. Mellentin JD, Murre C, Donlon TA et al. The gene for enhancer binding proteins E12/E47 lies at the t(1;19) breakpoint in acute leukemias. Science 1989; 246: 379–382. 40. Kaname T, Miyauchi T, Kuwano A et al. Mapping basigin (BSG), a member of the immunoglobulin superfamily, to 19p13.3. Cytogenet Cell Genet 1993; 64: 195–197. 41. Gozzetti A, Davis EM, Espinosa R et al. Identification of novel cryptic translocations involving IGH in B-cell non-Hodgkin’s lymphomas. Cancer Res 2002; 62: 5523–5527.

Downloaded from http://annonc.oxfordjournals.org/ at FU BerlinFB Humanmedizin on June 28, 2015

32. Ahmann GJ, Jalal SM, Juneau AL et al. A novel three-color, clonespecific fluorescence in situ hybridization procedure for monoclonal gammopathies. Cancer Genet Cytogenet 1998; 101: 7 –11. 33. Blade´ J, Samson D, Reece D et al. Criteria for evaluating disease response and progression in patients with multiple myeloma treated by high-dose therapy and haemopoietic stem cell transplantation. Br J Haematol 1998; 102: 1115–1123. 34. Huang SY, Tang JL, Yao M et al. Reduction of leukocyte count is associated with thalidomide response in treatment of multiple myeloma. Ann Hematol 2003; 82: 558–564. 35. Rajkumar SV, Fonseca R, Dewald GW et al. Cytogenetic abnormalities correlate with the plasma cell labeling index and extent of bone marrow involvement in myeloma. Cancer Genet Cytogenet 1999; 113: 73–77. 36. Kaufmann H, Kro¨mer E, No¨sslinger T et al. Both chromosome 13 abnormalities by metaphase cytogenetics and deletion of 13q by interphase FISH only are prognostically relevant in multiple myeloma. Eur J Haematol 2003; 71: 179–183.