Immunohistochemical Analysis Identifies Two Cyclin D1+ Subsets of Plasma Cell Myeloma, Each Associated With Favorable Survival

Transcription

1 Hematopathology / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA Immunohistochemical Analysis Identifies Two Cyclin D1+ Subsets of Plasma Cell Myeloma, Each Associated With Favorable Survival James R. Cook, MD, PhD, 1 Eric D. Hsi, MD, 1 Sarah Worley, 2 Raymond R. Tubbs, DO, 2 and Mohamad Hussein, MD 3 Key Words: Cyclin D1; Immunohistochemistry; Myeloma; Prognosis Abstract The significance of cyclin D1 expression in plasma cell myeloma was examined by immunohistochemical analysis using a newly available rabbit monoclonal antibody with superior staining properties. Two patterns of positive staining were observed, each associated with distinct pathologic features. Strong cyclin D1 staining was associated with increased plasma cells at diagnosis (P =.026), lymphoplasmacytic morphologic features (P =.029), CD20 expression (P =.040), and t(11;14)(q13;q32) as detected by interphase fluorescence in situ hybridization with simultaneous CD138 immunofluorescence (P <.001). In contrast, weak staining was associated with hyperdiploidy (P =.02) and gains of the CCND1 locus (P =.01). Overall survival was longer in the cyclin D1+ cases than in cyclin D1 cases (estimated 3-year survival, 73% vs 27%; P =.005). Improved survival was seen in the strongly positive and weakly positive groups compared with cyclin D1 cases (P =.005). This report shows for the first time that cyclin D1 immunohistochemical analysis can provide prognostic information in plasma cell myeloma. Plasma cell myeloma, also known as multiple myeloma, is a bone marrow based malignant neoplasm of plasma cells associated with serum and/or urine monoclonal paraproteins and lytic skeletal lesions. 1 The molecular pathogenesis of plasma cell myeloma is complex. Although only a minority of cases display abnormal karyotypes by conventional cytogenetic studies, more sensitive fluorescence in situ hybridization (FISH) assays demonstrate chromosomal abnormalities to be nearly universal in myeloma, including numeric and structural changes. 2,3 Approximately 50% of cases show translocations involving the immunoglobulin heavy chain (IGH) locus and one of several partner genes. 3,4 The most common of these is t(11;14)(q13;q32) involving the cyclin D1 gene (CCND1) and IGH, which is present in approximately 15% to 20% of cases. 2,3,5,6 Recent studies have suggested that dysregulation of at least 1 of the cyclin D genes (CCND1, CCND2, and/or CCND3) is a common, unifying pathogenic event in multiple myeloma. 5,7 Some abnormalities involving the cyclin D genes might be associated with differences in prognosis. For example, the presence of t(11;14)(q13;q32) was reported to be associated with a favorable prognosis in plasma cell myeloma treated with high-dose chemotherapy and bone marrow transplantation. 8 Increased levels of cyclin D1 as detected by quantitative reverse transcriptase polymerase chain reaction (RT-PCR), whether secondary to a translocation or due to other abnormalities, also have been reported to be associated with a favorable prognosis. 9 Other laboratories, however, have been unable to confirm the association of t(11;14)(q13;q32) with a favorable prognosis. 10,11 The clinical significance of cyclin D1 dysregulation in plasma cell myeloma, therefore, has been uncertain. Downloaded from Am J Clin Pathol 2006;125:

2 Cook et al / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA We used a newly available rabbit monoclonal cyclin D1 antibody with superior staining properties 12 to study cyclin D1 protein expression in plasma cell myeloma. In an initial series of 20 cases, we correlated the results of immunohistochemical analysis with conventional cytogenetic analysis and interphase FISH studies for t(11;14)(q13;q32) performed on intact paraffin sections of bone marrow clot preparations using simultaneous CD138 immunofluorescence. In a second set of 44 cases of newly diagnosed plasma cell myeloma in which patients were enrolled in a phase 2 study of melphalan, prednisone, and rituximab, we examined the association of cyclin D1 immunohistochemical staining patterns with other clinicopathologic findings at diagnosis and with progression-free and overall survival. The findings showed that immunohistochemical analysis can identify 2 cyclin D1+ subsets of myeloma that display distinct clinicopathologic features but share a favorable prognosis. Materials and Methods Patients All studies were approved by the institutional review board of the Cleveland Clinic Foundation, Cleveland, OH. For correlation of cyclin D1 immunohistochemical results with those of conventional cytogenetic studies and FISH for t(11;14)(q13;q32), 20 cases of multiple myeloma were identified from the case files of the Cleveland Clinic Foundation, including 8 newly diagnosed and 12 patients with relapsed or refractory disease. Cases were required to have plasma cells present in a B-5 fixed core biopsy specimen and a formalin-fixed, paraffin-embedded clot section. For analysis of clinicopathologic characteristics and prognosis in newly diagnosed multiple myeloma, we studied 45 patients enrolled in a phase 2 study designed to evaluate the role of rituximab in sensitizing myeloma cells to traditional therapy with melphalan and prednisone, as well as the effect of longterm rituximab maintenance after the completion of the melphalan and prednisone on progression-free survival. 13 Patients with newly diagnosed multiple myeloma were eligible for this study. Therapy with oral melphalan was initiated at the standard dose of 0.25 mg/kg for days 1 through 4, not to exceed 24 mg/d, and prednisone, 100 mg/d for 4 days. This cycle was repeated every 4 to 6 weeks depending on peripheral blood counts. Rituximab was administered at a dose of 375 mg/m 2 weekly to be repeated every 6 months for a total of 6 treatments. Patients achieving stable disease or better continued with therapy until disease progression. Following the completion of the rituximab period of therapy, patients were observed without any specific treatment. Progression of disease was defined as hypercalcemia, new bone lesions, or an increase in serum monoclonal paraprotein by 30% or more in 2 consecutive samples. Following progression, patients were removed from the study with continuing survival follow-up. Of the 45 enrolled patients, 1 was excluded from this analysis because archived diagnostic pathologic specimens were not available. Clinicopathologic Characteristics in Newly Diagnosed Plasma Cell Myeloma The initial diagnostic specimen obtained following protocol enrollment was reviewed, including all available aspirate smears, bone marrow core biopsy specimens, bone marrow clot sections, and immunohistochemical stains. Cases were classified by morphologic type using the criteria of Bartl et al. 14 Clinical data were reviewed including age, sex, β 2 - microglobulin level at the time of enrollment, and serum and/or urine protein electrophoresis studies. Results of 4-color flow cytometric immunophenotypic studies performed at the time of diagnosis were reviewed when available (41 patients). A monoclonal plasma cell population could not be detected by flow cytometry in 7 studies. In the remaining 34 cases, the expression of CD20 and CD56 on plasma cells was analyzed. Cases were classified as positive for an antigen when there was expression on 20% or more of plasma cells. Cytogenetics Conventional cytogenetic studies of the bone marrow were performed at the Cleveland Clinic Foundation (n = 6) or at an outside reference laboratory (Dynagene, Houston, TX) (n = 14). For cases performed at the Cleveland Clinic Foundation, samples were cultured for 24 hours without mitogens. Parallel cultures stimulated with granulocyte macrophage colony-stimulating factor, TPA (tetradecanoyl phorbol 12-myristate 13-acetate), and/or Conditioned Medium III (derived from human bladder carcinoma cell line 5637) also were used. In cases analyzed at the reference laboratory, 48-hour unstimulated or 72-hour cultures stimulated with lipopolysaccharide were used. Chromosomes were identified by trypsin-giemsa banding and classified according to the International System for Cytogenetic Nomenclature. 15 FISH With Simultaneous CD138 Immunofluorescence Interphase FISH studies were performed on 5-µm paraffin sections of bone marrow clot preparations using a dualcolor, dual-fusion probe for t(11;14)(q13;q32) (Vysis, Downers Grove, IL) and simultaneous CD138-immunofluorescence with tyramide signal amplification (TSA/Alexa 350 kit No. 7, Molecular Probes, Eugene, OR). Paraffin slides were baked at 60 C overnight. Slides were deparaffinized by 3 immersions in xylene for 5 minutes each, followed by two 1-minute immersions in 100% alcohol, two 1-minute immersions in 95% alcohol, and a final 5-minute immersion in molecular grade Milli-Q water (Millipore, Billerica, MA). 616 Am J Clin Pathol 2006;125: Downloaded 616 from

3 Hematopathology / ORIGINAL ARTICLE Slides were heated to 95 C for 40 minutes followed by 20 minutes of cooling at room temperature. Slides were rinsed 3 times for 5 minutes each in phosphate-buffered saline (PBS)/0.1% polysorbate 20. Next, 50 µl of 1:20 CD138 antibody (clone BB4, Serotec, Raleigh, NC) was added with incubation for 1 hour in the dark. Slides were washed 3 times in PBS/0.1% polysorbate 20. Blocking reagent was added (100 µl/slide) with incubation in a humidified box for 30 minutes at room temperature. Horseradish peroxidase solution, 100 µl, was added with a 30-minute incubation at room temperature in a humidified box. Slides were washed 3 times in PBS/1% polysorbate 20. Tyramide solution, 100 µl per slide, was added, followed by incubation for 10 minutes at room temperature in the dark. Slides were washed 3 times, and 150 µl of a 1:5,000 dilution of Proteinase K solution (1 µl Proteinase K in 5 ml of a 50-mmol/L concentration of tris(hydroxymethyl)aminomethane hydrochloride, ph 7.6) was added to the slide, with incubation at room temperature for 16 minutes. Slides were washed for 5 minutes in Milli-Q water, dehydrated in alcohol, and allowed to completely air dry. Next, 1 µl of probe was added to 2 µl of water and 7 µl hybridization buffer (Vysis) in a microfuge tube. The 10-µL probe solution was added to each slide and a coverslip sealed over the slide with rubber cement. The probe and target DNA were allowed to codenature at 73 C for 5 minutes and then were hybridized overnight at 37 C. Next, slides were soaked in 2 standard saline citrate (SSC; ph 7.0) for 5 minutes at room temperature, and coverslips were removed. Slides were soaked in 2 SSC/0.1% NP40 for 3 to 5 seconds, rinsed in 2 SSC and then water, and allowed to air dry completely in darkness. Signals were visualized on an Axioskop photomicroscope (Zeiss, Oberkochen, Germany). Plasma cells were identified specifically by cytoplasmic CD138 staining under a DAPI (4',6-diamidino-2-phenylindole) filter. In each case, 100 to 200 plasma cells were counted. Nuclei were scored as positive (at least 1 fusion signal present), gain of CCND1 (3 or more copies of a red CCND1 signal per nucleus), or negative (all other patterns). Cases were considered positive for t(11;14)(q13;q32) when more than 10% of plasma cells exhibited at least 1 fusion signal. Cases were considered positive for gains of CCND1 when more than 10% of plasma cells exhibited 3 or more CCND1 signals per nucleus. The threshold of 10% represents 2.5 SD above the mean for fusion signals observed in negative control specimens (n = 5). Immunohistochemical Analysis In the initial series of 20 cases, B-5 fixed bone marrow core biopsy specimens were analyzed by immunohistochemical analysis. In the subsequent series of 44 patients with newly diagnosed myeloma, the earliest diagnostic specimen following enrollment on the protocol was selected for analysis. This latter set consisted of B-5 fixed bone marrow core biopsy specimens (n = 39), formalin-fixed bone marrow clot sections (n = 3), and formalin-fixed, directed bone biopsy specimens (n = 2). Immunohistochemical staining was performed using an automated immunostainer (Ventana Medical Systems, Tucson, AZ). The anti cyclin D1 antibody (clone SP4, Lab Vision, Fremont, CA) was used at a 1:250 dilution, and the anti-cd138 antibody (clone BB4, Serotec) was used at a 1:200 dilution. Heat-induced antigen retrieval was accomplished using CC1 solution (Ventana). In each case, the number of plasma cells present was estimated by review of the CD138 immunostain. Subsequently, the cyclin D1 immunostain was reviewed and the percentage of cyclin D1+ plasma cells estimated. Cases were categorized as negative (no nuclear staining in plasma cells detected), weakly positive (<50% of plasma cell nuclei positive), or strongly positive ( 50% of plasma cell nuclei positive). Statistics The Kaplan-Meier method was used to estimate survival time and progression-free survival time within cyclin D1 status groups and high- and low-stage groups; the groups were compared for survival time using the log-rank test. A Cox proportional hazards regression model was used to compare cyclin D1 groups for survival time while adjusting for stage. In the analysis of clinicopathologic characteristics, cyclin D1 status groups were compared on categorical variables using the Fisher exact test and on continuous variables using the 2- sample t test. All tests were 2-tailed and performed at a significance level of.05. SAS 8.2 (SAS Institute, Cary, NC) and GraphPad Prizm software (GraphPad Software, San Diego, CA) were used. Results Correlation of Cyclin D1 Immunohistochemical Findings With Cytogenetic Abnormalities The expression of cyclin D1 protein was correlated with conventional and molecular cytogenetic analysis in 20 cases of plasma cell myeloma Table 1. Two patterns of cyclin D1 positivity were observed Image 1 : 5 cases (25%) showed strong nuclear staining, 9 cases (45%) showed weak staining, and 6 cases (30%) were negative for cyclin D1. FISH studies for t(11;14)(q13;q32) were performed using simultaneous immunofluorescence for CD138 to specifically identify plasma cells Image 2 and Image 3. Nuclei were examined for the presence of IGH/CCND1 fusion signals or gains of the CCND1 locus. Each of the 5 cases with strong staining for cyclin D1 by immunohistochemical analysis was positive for t(11;14)(q13;q32) by FISH, with fusion signals identified in Downloaded from Am J Clin Pathol 2006;125:

4 Cook et al / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA Table 1 Correlation of Cyclin D1 Immunohistochemical Results With Results of Conventional Cytogenetic Studies and FISH for t(11;14)(q13;q32) in 20 Cases of Plasma Cell Myeloma Cyclin D1 FISH FISH Staining t(11;14) * +CCND1 * Karyotype Strong + (90) 46,XY[20] Strong + (62) 46,XY[20] Strong + (86) 46,XX[19] Strong + (89) 46,XY[20] Strong + (93) 46,XX[20] Weak + (98) 46,XY[20] Weak 46,XX[20] Weak + (71) 46,XX[20] Weak + (57) 55,X, X,del(1)(p22p33),+3,+5,+7,+9,+9,+11,+15,+15,+21,+22[5]/46,XX[15] Weak + (87) 44-46,X, X,del(1)(p13p22), 2,+add(3)(p12),der(4)t(2;4)(q14.1;q35),del(5)(q31q33),add(6)(p23), 8,+9,del(11)(q14),+add(11)(q23),+14,add(14)(p11.2),der(14)t(X;14)(q13;p13), add(17)(q21),+18,i(18)(q10), 20, 21,add(22)(q11.2),+1-2mar[cp16]/46,XX[1] Weak + (64) 56-57,XX,add(6)(q15),+add(6),+7,+9,add(10)(q26),+11,+14,+15,+15,+19,+21,+22, +mar[cp8]/46,xx[12] Weak + (65) 50-52,X, Y[6],+3[6],add(4)(p16)[6],+7[6],add(8)(p23)[4],+9,+11[5],+15[5], 16[6],+19[6], +19[6], 21[6],+2 mar[6][cp6]/46,xy[14] Weak + (90) 52,X, X,add(2)(q37),+3,+5,t(8;9)(p21;q13),+9,+11,+15,+19,+21[2]/5153,idem, 19[2], 22[1],+1-3mar[3][cp3]/46,XX[15] Weak + (73) 49-50,XY,add(1)(p22),+3,der(5)t(5;22)(q3?3;q11.2),+7,dic(7;16)(q36;p13.3),del(8)(p21),+9, +11,+der(11)del(11)(q13q14),del(11)(q24),der(12;22)(q10;q10),+15,+15, 22[cp5]/46,XY[15] + (45) 53-54,X,+X, Y,dic(1;?)(p12;?),+3,+7,del(8)(p11.2p21),+9, 10,+11,+15,+add(16)(q22), add(18)(q23),+add(19)(q13.1),+21, 22,+1-2mar[cp3]/46,XY[17] 46,XY[20] + (92) 46,XY[20] 46,XX[20] + (84) 46,XY[20] 43-44,XY,+add(1)(p1),del(3)(p13p21), 4,del(10)(q2?3),add(12)(p11.2), 14,add(15)(p11.2), add(16)(q2),add(22)(q1),+mar[cp2]/46,xy[18] FISH, fluorescence in situ hybridization; +, positive;, negative. * Numbers in parentheses are percentage of CD138+ nuclei. 62% to 93% of plasma cell nuclei identified (median, 89%). None of the cases with weakly positive or negative staining for cyclin D1 by immunohistochemical analysis were positive for t(11;14)(q13;q32). The correlation between the strong pattern of nuclear staining and the presence of t(11;14)(q13;q32) was highly significant (P <.0001). In contrast, the weakly cyclin D1+ cases were associated significantly with a hyperdiploid karyotype by conventional cytogenetic studies (6 [67%] of 9 weakly positive cases vs 1 [9%] of 11 others; P =.02), and gains of the CCND1 locus by FISH (8 [89%] of 9 weakly positive cases vs 3 [27%] of 11 others; P =.01). When present, gains of CCND1 also were identified in a large percentage of plasma cell nuclei (range, 45%-98%; median, 73%). Cyclin D1 Staining Patterns and Clinicopathologic Characteristics at Initial Diagnosis To examine the relationship between the pattern of cyclin D1 staining and clinicopathologic characteristics at initial diagnosis, cyclin D1 immunohistochemical analysis next was performed in a separate series of 44 patients with newly diagnosed plasma cell myeloma. Positive staining for cyclin D1 in plasma cells was identified in 22 (50%) of 44 cases. In 8 cases (18%), expression was classified as strong, whereas 14 cases (32%) showed weak expression. The remaining 22 cases (50%) showed no evidence of cyclin D1+ plasma cells. The clinicopathologic characteristics of the cyclin D1+ and cyclin D1 cases are shown in Table 2. Comparison of cyclin D1 cases with cases showing any positivity for cyclin D1 showed no significant differences in age, sex, initial β 2 -microglobulin levels, number of bone marrow plasma cells, plasma cell immunophenotype, or the type of associated monoclonal paraprotein. However, separate analysis of the strongly positive, weakly positive, and negative subsets demonstrated that strong cyclin D1 staining was associated significantly with increased numbers of bone marrow plasma cells at diagnosis (P =.026), small cell (lymphoplasmacytic) morphologic type (P =.029), and coexpression of CD20 (P =.040). Cyclin D1 Staining Patterns and Prognosis in Newly Diagnosed Myeloma The progression-free and overall survival times for cyclin D1+ and cyclin D1 cases were compared. As shown in Figure 1A, the cyclin D1+ cases showed longer progression-free survival, although the difference did not reach statistical significance (median, 15.7 months for cyclin D1+ vs 12.8 months for 618 Am J Clin Pathol 2006;125: Downloaded 618 from

5 Hematopathology / ORIGINAL ARTICLE A B C Image 1 Patterns of cyclin D1 immunohistochemical staining in plasma cell myeloma. A, In some cases, no nuclear staining for cyclin D1 was observed in plasma cells ( 400). B, Example of weakly positive staining for cyclin D1 (<50% of plasma cells showing positive nuclear staining). The intensity of staining in weakly positive cases was somewhat variable but generally was quite weak and visible only at high power ( 400). C, A subset of cases demonstrated strong nuclear staining for cyclin D1 ( 50% plasma cells with nuclear staining). The intensity of staining in strongly positive cases was characteristically very intense, and cytoplasmic staining also was present ( 400). cyclin D1 ; P =.13). However, the overall survival was significantly longer in cyclin D1+ cases than in cyclin D1 cases Figure 1B. The estimated 3-year survival was 73% in the cyclin D1+ subset compared with 27% in the cyclin D1 subset (P =.001). In contrast, CD20 expression was not associated with a significant difference in overall survival (estimated 3- year survival 60% for CD20+ vs 41% for CD20 ; P =.19). The relationship between stage at presentation, cyclin D1 status, and overall survival time was studied. A similar proportion of cases presented at high stage (Southwest Oncology Group stage 3 or 4) in the cyclin D1+ (2/22 [9%]) and cyclin D1 (5/22 [23%]) subsets (P =.41). Low-stage cases showed a slightly improved estimated 3-year overall survival compared with high-stage cases (54% vs 29%), although this difference was not statistically significant (P =.40). In multivariable analysis, cyclin D1 positivity remained associated with a prolonged overall survival time after adjustment for stage (P =.003). To determine whether the improved survival time in cyclin D1+ cases was limited to the strongly positive cases, we compared progression-free and overall survival times in 3 groups: cyclin D1, cyclin D1 weakly positive, and cyclin D1 strongly positive. The median progression-free survival time was 12.8 months in the negative group, 15.7 months in the weakly positive group, and 21.4 months in the strongly positive group (P =.31) Figure 2A. As shown in Figure 2B, prolonged overall survival was seen in the weakly and strongly positive groups compared with the cyclin D1 subset. The estimated 3-year survival rates in cyclin D1, weakly positive, and strongly positive groups were 27%, 71%, and 75%, respectively (P =.005). In multivariable analysis, this difference remained significant after adjustment for stage (P =.011). Comparisons between groups showed that overall survival was significantly longer in the strongly positive vs negative cases (P =.038) and in the weakly positive vs negative cases (P =.006). Downloaded from Am J Clin Pathol 2006;125:

6 Cook et al / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA Image 2 Fluorescence in situ hybridization analysis for t(11;14)(q13;q32) with simultaneous CD138 immunofluorescence. Under a DAPI (4',6-diamidino-2- phenylindole) filter, CD138+ plasma cells are identified by intense blue cytoplasmic staining, whereas other marrow elements remain unlabeled. Only nuclei of CD138+ plasma cells are scored. A B C Image 3 Fluorescence in situ hybridization analysis for t(11;14)(q13;q32) in plasma cell myeloma. A, In normal cells, 2 red signals (CCND1) and 2 green signals (immunoglobulin heavy chain [IGH]) are observed. B, A gain of CCND1 is identified as 3 or more red CCND1 signals. C, The presence of t(11;14)(q13;q32) is identified in this case showing 1 red CCND1 signal, 1 green IGH signal, and 2 yellow fusion signals representing the derivative chromosomes. 620 Am J Clin Pathol 2006;125: Downloaded 620 from

7 Hematopathology / ORIGINAL ARTICLE Table 2 Clinicopathologic Characteristics of Subsets of Plasma Cell Myeloma Positive and Negative for Cyclin D1 * Negative Positive P Strongly Positive P Weakly Positive P Age (y) 59.7 ± ± ± ± Male 14/22 (64) 14/22 (64).99 4/8 (50).68 10/14 (71).73 β 2 m (mg/l) 3.5 ± ± ± ± No. of BM PCs (%) 45.5 ± ± ± ± Morphologic type (Bartl et al 14 ) Marschalko 13/22 (59) 8/22 (36).23 1/8 (13).040 7/14 (50).73 Asynchronous 6/22 (27) 6/22 (27).99 1/8 (13).64 5/14 (36).72 Small cell 2/22 (9) 6/22 (27).24 4/8 (50).029 2/14 (14).63 Cleaved 1/22 (5) 1/22 (5).99 1/8 (13).47 0/14 (0).99 Blastic 0/22 (0) 1/22 (5).99 1/8 (13).27 0/14 (0) Phenotype CD20+ 1/17 (6) 4/17 (24).33 3/6 (50).040 1/11 (9).99 CD56+ 14/17 (82) 15/17 (88).99 5/6 (83).99 10/11 (91).99 Paraprotein IgG 8/22 (36) 11/22 (50).54 3/8 (38).99 8/14 (57).31 IgA 6/22 (27) 4/22 (18).72 1/8 (13).64 3/14 (21).99 Light chain only 8/22 (36) 7/22 (32).99 4/8 (50).68 3/14 (21).47 κ 14/22 (64) 16/22 (73).75 5/8 (63).99 11/14 (79).47 β 2 m, β 2 -microglobulin level; BM, bone marrow; PCs, plasma cells. * Data are given as number/total (percentage) or mean ± SD. Significant P values are given in bold type. β 2 m values are given in conventional units; to convert to Système International units (nmol/l), multiply by Positive vs negative. Strongly positive vs negative. Weakly positive vs negative. A B % Progression-Free Survival Cyclin D1 Negative Positive Start Time (y) % Survival Cyclin D1 Negative Positive Start Time (y) Figure 1 Cyclin D1 expression is associated with a favorable prognosis in plasma cell myeloma. A, Cyclin D1+ cases showed longer progression-free survival, although this difference was not statistically significant (P =.13). B, Overall survival was significantly longer in cyclin D1+ cases (P =.001). There was no difference in overall survival between cases showing strongly vs weakly positive staining (P =.31). Discussion Cyclin D1 protein is expressed aberrantly in a subset of plasma cell myeloma owing to one of several genetic abnormalities. 3-6,16 Approximately 15% to 20% of cases are associated with t(11;14)(q13;q32), which juxtaposes CCND1 with immunoglobulin heavy chain enhancer elements, leading to expression of cyclin D1 protein. 3,4,17 Cyclin D1 may also be expressed in association with polysomy for chromosome 11, 9,18,19 amplification of CCND1, 20 or other abnormalities. 5 Studies of the prognostic significance of t(11;14)(q13;q32) in plasma cell myeloma have yielded conflicting results. Moreau et al 8 reported that the presence of t(11;14)(q13;q32) was associated with a markedly prolonged overall survival in patients treated with high-dose chemotherapy and bone marrow Downloaded from Am J Clin Pathol 2006;125:

8 Cook et al / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA A B % Progression-Free Survival Cyclin D1 Negative Weak positive Strong positive Start Time (y) % Survival Cyclin D1 Negative Weak positive Strong positive Start Time (y) Figure 2 Cyclin D1 strongly positive and weakly positive subsets show favorable overall survival. A, When analyzed in 3 groups (cyclin D1 strongly positive, weakly positive, and negative) a significant difference in progression-free survival was not identified (P =.31). B, The strongly positive and weakly positive subsets showed longer overall survival than the cyclin D1 subset (P =.005). transplantation, although another series of patients with apparently similar therapy showed no effect of this translocation on prognosis. 11 A study of patients treated with conventional chemotherapy showed a trend toward better survival in t(11;14)(q13;q32)-positive cases, but this difference did not reach statistical significance. 21 More recently, the prognostic significance of cyclin D1 dysregulation in general, rather than only cyclin D1 translocations, has been studied. Quantitative RT-PCR studies showed that increased cyclin D1 expression, including cases due to either t(11;14)(q13;q32) or other abnormalities, is associated with a longer duration of remission in patients treated with high-dose chemotherapy and bone marrow transplantation. 9 Considered together, the previous findings suggest that an assessment of cyclin D1 status in plasma cell myeloma might provide clinically useful prognostic information. However, FISH studies and quantitative RT-PCR are complex techniques not available in many routine pathology laboratories. In addition, plasma cells in many cases might represent only a minority of the bone marrow cellularity, such that before analysis, purification of plasma cells from a bone marrow sample may be necessary using cell sorting or other immunoselective techniques. For these reasons, we studied the significance of cyclin D1 abnormalities in myeloma using immunohistochemical analysis, which does not require cell sorting and can be used in most routine pathology laboratories. Previous studies also have examined cyclin D1 protein expression in myeloma using immunohistochemical analysis and have reported cyclin D1 positivity in 24% to 26% of cases. 19,22-24 Although t(11;14)(q13;q32)-positive cases typically are positive for cyclin D1 protein by immunohistochemical analysis, only a subset of cases with detectable cyclin D1 staining is associated with t(11;14)(q13;q32). 19,24,25 Previous studies, however, have been limited by the nature of most available cyclin D1 antibodies, which generally are associated with high background and variable staining intensities in most laboratories. A rabbit monoclonal antibody against cyclin D1 has become available that shows superior staining properties and lower background staining compared with murine antibodies. 12 In the present study, by using this rabbit monoclonal antibody, we found evidence of cyclin D1 expression in 22 (50%) of 44 cases of newly diagnosed myeloma. This incidence is similar to the results of RT-PCR based studies that found overexpression of cyclin D1 messenger RNA in 43% 9 and 50% 26 of cases. Two patterns of cyclin D1 staining were identified. The strongly positive cases were uniformly positive for t(11;14)(q13;q32) by FISH and were associated with increased numbers of plasma cells (P =.026), lymphoplasmacytic morphologic type (P =.029), and CD20 expression (P =.040). Previous studies also have shown that lymphoplasmacytic morphologic type and CD20 expression are associated with the presence of the t(11;14)(q13;q32). 27,28 In contrast, the weakly positive cases uniformly lacked the t(11;14)(q13;q32) but were associated with a hyperdiploid karyotype by conventional cytogenetic studies (P =.02) and gains of the CCND1 locus by FISH (P =.01). It is interesting that these results are similar to those described in gene expression profiling experiments that described one subset of multiple myeloma with high levels of cyclin D1 expression secondary to t(11;14)(q13;q32) and a second subset with lower levels of cyclin D1 that lacked t(11;14)(q13;q32) but frequently was 622 Am J Clin Pathol 2006;125: Downloaded 622 from

9 Hematopathology / ORIGINAL ARTICLE hyperdiploid. 7,16 These findings show that immunohistochemical analysis can be used to identify subsets of plasma cell myeloma analogous to those initially defined by gene expression profiling data. In the series of 20 cases examined in the present study using immunohistochemical analysis and interphase FISH analysis, the strong pattern of cyclin D1 staining showed 100% sensitivity and specificity for the presence of t(11;14)(q13;q32). This finding suggests that immunohistochemical analysis could be used as a surrogate for FISH studies to demonstrate the presence of this translocation. However, the number of translocation-positive cases examined remains small, and additional studies are warranted to more thoroughly compare the sensitivity of immunohistochemical analysis with FISH for the detection of t(11;14)(q13;q32). It is important to note that this study demonstrates that cyclin D1+ cases of myeloma display a significantly longer overall survival time. The time to progression also was longer in cyclin D1+ cases, although this finding did not reach statistical significance. These data show, for the first time to our knowledge, that immunohistochemical staining for cyclin D1 can provide important prognostic information in plasma cell myeloma. In contrast, a previous study of cyclin D1 immunohistochemical analysis in 59 cases of myeloma showed no differences in outcome between positive and negative cases. 23 This discrepancy might be due to differences in the antibodies used and the smaller percentage of cyclin D1+ cases identified in the earlier analysis. The present study confirmed the finding of the RT- PCR based study of Soverini et al 9 that increased cyclin D1 expression is a favorable prognostic factor in plasma cell myeloma. Furthermore, our findings provide the first evidence that the improved survival in cyclin D1+ cases is not limited to therapy regimens that include bone marrow transplantation. In additional analysis, prolonged overall survival times were seen in the cyclin D1 strongly positive and weakly positive subsets. This observation suggests that the favorable survival in these cases is not related to the presence of t(11;14)(q13;32) per se but instead is seen in association with any detectable cyclin D1 protein expression. This result is similar to that described in gene expression profiling experiments in which an improved prognosis was said to be associated with subsets of cases with high levels of cyclin D1 due to t(11;14)(q13;q32) or low-level cyclin D1 expression associated with trisomy Considered together, these results suggests that it may be more useful in routine practice to assess for the presence of any cyclin D1 expression, regardless of the underlying cytogenetic abnormality, rather than identifying only the presence or absence of t(11;14)(q13;q32). Additional studies will be required to directly compare these types of analyses. Recently, a molecular classification of plasma cell myeloma has been proposed based upon results of gene expression profiling. This system, designated the TC classification, separates cases based on the presence of specific recurring translocations involving the IGH locus and the expression of cyclin D1, cyclin D2, and/or cyclin D3. 5,16 As discussed, our findings indicate that immunohistochemical analysis can be used to identify at least some subsets of myeloma analogous to those found by expression microarray analysis. In a similar manner, others have shown that subtypes of diffuse large B-cell lymphoma initially defined by gene expression profiling can be identified appropriately by using a series of immunohistochemical stains. 30 Additional studies are ongoing to determine whether immunohistochemical analysis for cyclins D2 and D3 also can be used to identify additional subsets of patients with distinct clinical or pathologic features, analogous to those proposed in the TC classification. Immunohistochemical analysis for cyclin D1 can identify 2 distinct subsets of plasma cell myeloma. Although the strongly positive and weakly positive subsets display different pathologic features, both subsets share a favorable overall survival. The favorable effect of cyclin D1 expression, therefore, can be demonstrated by immunohistochemical staining rather than RT-PCR, FISH, or other molecular techniques. Because prognostic markers might show variable effects in different therapeutic regimens, these findings should be confirmed in patients receiving other types of therapy. Cyclin D1 immunohistochemical analysis is a straightforward, widely available technique that could be used in most routine pathology laboratories for prognostic assessment of plasma cell myeloma. From the 1 Department of Clinical Pathology, 2 Department of Quantitative Health Sciences, and 3 Multiple Myeloma Program, Cleveland Clinic Foundation, Cleveland, OH. Address reprint requests to Dr Cook: Dept of Clinical Pathology/L11, Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH References 1. Grogan TM, Van Camp B, Kyle RA, et al. Plasma cell neoplasms. In: Jaffe ES, Harris NL, Stein H, et al, eds. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001: World Health Organization Classification of Tumours. 2. Zandecki M, Lai JL, Facon T. Multiple myeloma: almost all patients are cytogenetically abnormal. Br J Haematol. 1996;94: Fonseca R, Barlogie B, Bataille R, et al. Genetics and cytogenetics of multiple myeloma: a workshop report. Cancer Res. 2004;64: Bergsagel PL, Kuehl WM. Chromosome translocations in multiple myeloma. Oncogene. 2001;20: Bergsagel PL, Kuehl WM. Critical roles for immunoglobulin translocations and cyclin D dysregulation in multiple myeloma. Immunol Rev. 2003;194: Hideshima T, Bergsagel PL, Kuehl WM, et al. Advances in biology of multiple myeloma: clinical applications. Blood. 2004;104: Downloaded from Am J Clin Pathol 2006;125:

10 Cook et al / CYCLIN D1 IMMUNOHISTOCHEMISTRY IN MYELOMA 7. Bergsagel PL, Kuehl WM, Zhan F, et al. Cyclin D dysregulation: an early and unifying pathogenic event in multiple myeloma. Blood. 2005;106: Moreau P, Facon T, Leleu X, et al. Recurrent 14q32 translocations determine the prognosis of multiple myeloma, especially in patients receiving intensive chemotherapy. Blood. 2002;100: Soverini S, Cavo M, Cellini C, et al. Cyclin D1 overexpression is a favorable prognostic variable for newly diagnosed multiple myeloma patients treated with high-dose chemotherapy and single or double autologous transplantation. Blood. 2003;102: Fonseca R, Blood E, Rue M, et al. Clinical and biologic implications of recurrent genomic aberrations in myeloma. Blood. 2003;101: 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: Cheuk W, Wong KO, Wong CS, et al. Consistent immunostaining for cyclin D1 can be achieved on a routine basis using a newly available rabbit monoclonal antibody. Am J Surg Pathol. 2004;28: Hussein M, Karam M, McLain D, et al. Biologic and clinical evaluation of Rituxan (RT) in the management of newly diagnosed multiple myeloma (MM) patients (pts) [abstract]. Blood. 1999;94(suppl 1; pt 1): Bartl R, Frisch B, Fateh-Moghadam A, et al. Histologic classification and staging of multiple myeloma; a retrospective and prospective study of 674 cases. Am J Clin Pathol. 1987;87: Mitelman F, ed. ISCN 1995: An International System for Human Cytogenetic Nomenclature Basel, Switzerland: S Karger; Barille-Nion S, Barlogie B, Bataille R, et al. Advances in biology and therapy of multiple myeloma. Hematology (Am Soc Hematol Educ Program). 2003: 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: Specht K, Haralambieva E, Bink K, et al. Different mechanisms of cyclin D1 overexpression in multiple myeloma revealed by fluorescence in situ hybridization and quantitative analysis of mrna levels. Blood. 2004;104: Pruneri G, Fabris S, Baldini L, et al. Immunohistochemical analysis of cyclin D1 shows deregulated expression in multiple myeloma with the t(11;14). Am J Pathol. 2000;156: Hoechtlen-Vollmar W, Menzel G, Bartl R, et al. Amplification of cyclin D1 gene in multiple myeloma: clinical and prognostic relevance. Br J Haematol. 2000;109: Fonseca R, Blood EA, Oken MM, et al. Myeloma and the t(11;14)(q13;q32): evidence for a biologically defined unique subset of patients. Blood. 2002;99: Athanasiou E, Kaloutsi V, Kotoula V, et al. Cyclin D1 overexpression in multiple myeloma: a morphologic, immunohistochemical, and in situ hybridization study of 71 paraffin-embedded bone marrow biopsy specimens. Am J Clin Pathol. 2001;116: Markovic O, Marisavljevic D, Cemerikic V, et al. Immunohistochemical analysis of cyclin D1 and p53 in multiple myeloma: relationship to proliferative activity and prognostic significance. Med Oncol. 2004;21: Vasef MA, Medeiros LJ, Yospur LS, et al. Cyclin D1 protein in multiple myeloma and plasmacytoma: an immunohistochemical study using fixed, paraffin-embedded tissue sections. Mod Pathol. 1997;10: Hoyer JD, Hanson CA, Fonseca R, et al. The (11;14)(q13;q32) translocation in multiple myeloma: a morphologic and immunohistochemical study. Am J Clin Pathol. 2000;113: Troussard X, Avet-Loiseau H, Macro M, et al. Cyclin D1 expression in patients with multiple myeloma. Hematol J. 2000;1: Robillard N, Avet-Loiseau H, Garand R, et al. CD20 is associated with a small mature plasma cell morphology and t(11;14) in multiple myeloma. Blood. 2003;102: 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: Harousseau JL, Shaughnessy J Jr, Richardson P. Multiple myeloma. Hematology (Am Soc Hematol Educ Program). 2004: Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood. 2004;103: Am J Clin Pathol 2006;125: Downloaded 624 from