Anaplastic Large Cell Lymphoma A Flow Cytometric Analysis of 29 Cases

Transcription

1 Hematopathology / FLOW CYTOMETRY OF ANAPLASTIC LYMPHOMA Anaplastic Large Cell Lymphoma A Flow Cytometric Analysis of 29 Cases Melissa V. Kesler, MD, Geeta S. Paranjape, MD, Sheryl L. Asplund, MD, Robert W. McKenna, MD, Saba Jamal, MD, and Steven H. Kroft, MD Key Words: Flow cytometry; Anaplastic large cell lymphoma; Anaplastic lymphoma kinase; ALK Abstract We report our experience with flow cytometric (FC) analysis of 29 cases of anaplastic large cell lymphoma (ALCL). Morphologic analysis of processed cytocentrifuged preparations demonstrated neoplastic cells in 28 cases. In 25 of these, an aberrant lymphoid population was detected by FC analysis. The majority showed high orthogonal light scatter, similar to monocytes or granulocytes. Of the antigens CD2, CD3, CD4, CD5, and CD7, 5 cases expressed 1, 8 expressed 2, 6 expressed 3, 3 expressed 4, and 3 expressed all 5. CD4 was expressed most commonly (20/25 [80%]), followed by CD2 (18/25 [72%]), CD3 (10/25 [40%]), and CD5 and CD7 (8/25 [32%] each). CD45 was expressed in 23 of 25 cases and CD13 in 7 of 9. Of 21 cases, 13 were anaplastic lymphoma kinase (ALK)+, all of which were CD4+, vs 5 of 8 ALK cases (P =.042). Most ALCLs can be detected and characterized by multiparameter FC analysis. However, light scatter gating on typical lymphoid regions may yield falsenegative results in a substantial number of cases. Systemic anaplastic large cell lymphoma (ALCL) accounts for 3% of adult non-hodgkin lymphomas (NHLs) and 10% to 30% of childhood NHLs and, in general, is associated with a good response to therapy and a favorable outcome. Approximately 60% of ALCLs carry a translocation involving the anaplastic lymphoma kinase (ALK) gene on chromosome 2p23, most commonly a t(2;5)(p23;q35) involving the nucleophosmin (NPM) gene. 1,2 These rearrangements result in overexpression of ALK fusion proteins, and their presence is associated with an excellent prognosis. 3-6 Although ALCL may have a T-cell or null-cell immunophenotype by immunohistochemical analysis, all cases are considered genotypically of T-cell origin based on results of T-cell receptor gene rearrangement studies. 7 Extensive immunohistochemical data regarding the immunophenotype of ALCL are available in the literature. With the exception of the small cell variant, the majority of tumor cells in ALCL are by definition CD30+, and multiple studies have demonstrated variable positivity for a number of T cell and non T cell associated antigens, including CD2, CD3, CD4, CD8, CD25, CD43, CD45, CD45RO, CD56, CD13, CD15, CD33, HLA-DR, TIA-1, ZAP-70, and T-cell receptor molecules By comparison, relatively few studies have addressed the use of flow cytometric (FC) analysis in the evaluation of ALCL Furthermore, although it has been reported that 11% to 15% of NHLs in general and up to 25% of large cell lymphomas are not detected by FC analysis (for a variety of technical reasons), 21 few or no data are available regarding the efficiency of detection of ALCL by FC analysis. Finally, although the presence or absence of ALK expression by immunohistochemical analysis clearly divides ALCL into Am J Clin Pathol 2007;128: Downloaded 314 from

2 Hematopathology / ORIGINAL ARTICLE distinct clinicopathologic groups, the immunophenotypic relationship between ALK+ and ALK cases has not been well characterized. We surveyed our single-institution experience with flow cytometric analysis of ALCL for a nearly 10-year period with the aims of better characterizing the immunophenotypic features of this neoplasm, comparing antigen expression in ALK+ and ALK cases and assessing the sensitivity of FC analysis for the detection of ALCL in routine practice. Materials and Methods Case Selection The FC database at the University of Texas Southwestern, Dallas, was searched for all NHL cases between October 1994 and April 2004 for which the FC interpretation was ALCL. In addition, the institutional hematopathology database was searched for all cases diagnosed as ALCL, and cases were cross-referenced with the FC database. We retrieved 31 cases. After review of available morphologic and clinical data, 2 cases were excluded. One arose in a patient with chronic lymphocytic leukemia (CLL) and was not thought to be morphologically diagnostic of ALCL; instead, it was thought to represent a Hodgkin-like transformation of CLL. This case expressed CD30 and CD45 and lacked all B and T markers except CD8. The second excluded case was a primary cutaneous ALCL in a patient with underlying CLL. Thus, a total of 29 cases of systemic ALCL with FC immunophenotyping were the subject of this analysis. FC Analysis Tissue processing and antibody staining were performed as described previously. 22 Briefly, fresh biopsy tissue samples were sliced and disaggregated through a mesh smaller than 100 µm, and cells were suspended in 5% newborn calf serum in RPMI 1640 tissue culture medium (Life Technology, Rockville, MD). Cell counts were performed manually, and 500,000 cells per tube were washed with a phosphate-buffered saline/0.0455% sodium azide/0.1% bovine serum albumin solution (PAB) and then incubated with a 3- or 4-color combination of antibodies. Specimens were incubated at 2 C to 8 C in the dark for 20 minutes, washed with PAB, and resuspended in 1% paraformaldehyde in phosphate-buffered saline. Antibodies used in all cases included CD2 (55.2), CD3 (SK7), CD5 (L17F12), CD7 (4H9), CD4 (SK3), CD8 (SK1), CD45 (2D1), CD38 (HB7), CD19 (SJ25C1), CD20 (L27), and monoclonal κ (TB28-2) and λ (I155-2), Becton Dickinson, San Jose, CA; CD23 (B6), FMC7, and polyclonal goat κ and λ, Coulter-Immunotech, Hialeah, FL; and CD30 (Ber-H2), DAKO, Carpinteria, CA. Additional antibodies assessed in a subset of cases were CD45RO (UCHL-1, 17 cases), CD56 (MY31, 10 cases), CD15 (MMA, 4 cases), CD36 (FA6.152, 2 cases), and CD13 (L138, 9 cases), Becton Dickinson. Antibodies were conjugated with fluorescein isothiocyanate, phycoerythrin, peridinin chlorophyll protein, or allophycocyanin. A portion of each specimen was stained with isotype-matched fluorescent control monoclonal antibodies. Data were acquired using a 3-color FACScan or 4- color FACSCalibur with CELLQuest software (Becton Dickinson), and cluster analysis was performed on ungated data using Paint-a-Gate software (Becton Dickinson). FC Interpretation Distinct cell populations (clusters) were identified based on any combination of forward and orthogonal light scatter properties and fluorescence intensity with various antibody combinations. Nonviable cells and debris were excluded based on light scatter. An antigen was considered positive if expressed by at least 10% of ALCL events beyond a 2% threshold determined on the same population (based on light scatter) in the isotypic control tube. Bright and dim expression of T-cell antigens and CD45 were assessed relative to expression on residual normal T cells in the samples and also based on accumulated laboratory experience with individual antibodies in normal and neoplastic populations. 23 Histologic and Immunohistochemical Analyses Biopsy specimens were fixed in B-5 fixative and/or 10% neutral buffered formalin, embedded in paraffin, and sectioned at 3 µm. Sections were stained with H&E for light microscopic examination. Immunohistochemical staining for CD30 (Ber- H2, dilution 1:20; DAKO) and ALK (Alk-1, dilution 1:10; DAKO) was performed on paraffin sections in selected cases using a TechMate automated immunostainer (Ventana Biotek Systems, Tucson, AZ) with a standard streptavidin-biotin peroxidase detection system. For antigen retrieval, sections were placed in 200 ml of antigen-retrieval citrate buffer (BioGenex, San Ramon, CA), ph 6.8, with boiling for 5 minutes. After adding 50 ml of deionized water, the buffer was again brought to a boil for 5 minutes. The slides were allowed to cool in buffer for 20 minutes before further processing. Results Patient Information Specimens were from 17 males and 12 females (M/F ratio, 1.4:1) ranging in age from 5 to 79 years (median, 30 years). Sources of 26 specimens were nodal or soft tissue masses, 2 were from pathologically involved bone marrow, and 1 was from pleural fluid. One female patient was 40 weeks pregnant at the time of lymph node biopsy. Tissue sections were available for review in 27 cases. Of these, 23 Downloaded from Am J Clin Pathol 2007;128:

3 Kesler et al / FLOW CYTOMETRY OF ANAPLASTIC LYMPHOMA were common ALCL, 1 was a lymphohistiocytic variant, 2 were small cell variants, and 1 was mixed common and small cell variant. The 2 cases for which sections were unavailable were a staging bone marrow in a patient recently diagnosed with ALCL at an outside institution and a pleural fluid specimen. In both of these cases, the morphologic features of the cytocentrifuged preparations were typical for ALCL. Comparison of Cytocentrifuged Preparation and FC Results Wright-stained, postprocessing cytocentrifuged preparations were assessed in all cases for the presence of morphologically malignant cells. In 28 of 29 cases, malignant cells were identified in the cytocentrifuged preparation; 25 of these 28 yielded an aberrant population on FC analysis. In 1 of 3 cytocentrifuged cases that was positive but had negative FC results, few tumor cells were present on the cytocentrifuged preparations. The other 2 cases had negative FC results despite relatively abundant tumor cells by morphologic examination. In the remaining case (one of the small cell variants), the cytocentrifuged preparation was negative, as were the FC results. Review of direct tissue imprints revealed rare malignant cells. CD3-APC CD2-PE FC Findings in 25 Positive Cases Light Scatter Characteristics In 8 cases, the malignant cells fell in the large lymphocyte/blast region Image 1, whereas 1 case fell in the medium-sized lymphocyte/blast region Image 2. The remaining 16 cases had high orthogonal light scatter, similar to monocytes or granulocytes, with variable forward scatter Image 3, Image 4, Image 5, and Image 6. This finding typically (but not uniformly) correlated with the presence of prominent cytoplasmic vacuolization. Cases with high orthogonal scatter also often failed to form tight light scatter clusters, instead manifesting a relatively diffuse scatter pattern (Images 4-6). However, they formed tight clusters in one or more antibody tubes, allowing immunophenotypic detection. Antigen Expression The tumor cells expressed CD45 in 23 of 25 cases. In 18, expression was of normal intensity, whereas 5 cases underexpressed CD45 compared with normal lymphocytes. CD30 was positive in 23 of 25 cases. The remaining 2 cases were positive for CD30 by immunohistochemical analysis but negative by FC analysis. CD38 was expressed in 12 of 25 cases. All ALCL cases were immunophenotypically aberrant with respect to expression of the T-cell antigens CD2, CD3, CD4, CD5, and CD7. The number of aberrancies ranged from 1 to 5 per case (average, 3.9; median, 4), with 1 aberrancy in 1 case, 2 in 2 cases, 3 in 2 cases, 4 in 14 cases, and 5 in 6 cases. With respect to the T-cell antigens CD2, CD3, CD4, CD5, and CD19-APC CD5-PerCP CD7, 5 cases (20%) expressed only one of these (CD2 and CD4 in 2 cases each and CD7 in 1 case), 8 cases (32%) expressed 2, 6 cases (24%) expressed 3, 3 cases (12%) expressed 4, and 3 (12%) expressed all 5 antigens. CD4 was expressed most commonly (80%), followed by CD2 (72%), CD3 (40%), and CD5 and CD7 (32% each). T-cell antigens were often expressed at an abnormal intensity compared with normal T cells, dim in 48% and bright in 14% of cases Table 1. No case expressed CD8. CD45RO was expressed in 16 of 17 cases analyzed. CD56 was expressed in 6 of 10, HLA-DR in 6 of 6, CD25 in 2 of 3, CD15 in 2 of 4, CD36 in 1 of 2, and CD13 in 7 of 9 cases analyzed. B-cell markers (CD19, CD20, CD23, FMC7, and surface immunoglobulin) were negative in all cases. IgG1-PE IgG2a-FITC Image 1 (Case 15) This lymphohistiocytic, anaplastic lymphoma kinase positive anaplastic large cell lymphoma from a neck lymph node in a 62-year-old man demonstrates neoplastic cells (red) that scatter in the region one commonly encounters large cell lymphomas. The cells are CD30+/ CD45+/CD4+ but lack CD2, CD3, CD5, and CD7. Green, normal T cells. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. 316 Am J Clin Pathol 2007;128: Downloaded 316 from

4 Hematopathology / ORIGINAL ARTICLE CD2-PE CD8-PerCP CD19-APC CD3-APC CD3-APC CD2-PE CD8-PerCP CD13-PE CD5-FITC CD5-PerCP CD56-PE Image 2 (Case 10) This anaplastic lymphoma kinase positive anaplastic large cell lymphoma (red) from a cervical lymph node in an 18-year-old woman demonstrates light scatter characteristics in the medium-sized lymphocyte/blast region, corresponding to the unusually small neoplastic cells in tissue sections. The lymphoma cells express CD45, CD2, CD3, CD4, and CD5 at normal intensity, but CD7 is abnormally bright compared with normal lymphocytes (green). An interesting feature of this case is that it was CD30 by flow cytometry yet was CD30+ by immunohistochemical analysis. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. Image 3 (Case 20) This anaplastic large cell lymphoma (anaplastic lymphoma kinase status unknown) (red) from the bone marrow of a 30-year-old man demonstrates high orthogonal scatter, similar to granulocytes or monocytes. This was attributed to prominent cytoplasmic vacuolization of the neoplastic cells. The cells are CD2+/CD4+ but lack CD3 and CD5. They were also CD45+/CD7 (not shown). Finally, the neoplastic cells are CD56+/CD13bright+. Green, normal T cells. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. Comparison of ALK+ and ALK Cases Immunohistochemical staining for ALK protein was available in 21 of 25 cases; 13 were ALK+ and 8 were ALK Table 2. All ALK+ cases were CD4+, whereas 5 (63%) of 8 ALK cases expressed CD4 (P =.042). Furthermore, no ALK case expressed normal levels of CD4, compared with 6 (46%) of 13 ALK+ cases (P =.046). Of the 5 CD13+ cases for which ALK staining results were available, all were ALK+. Both CD13 cases were ALK. Both CD15+ cases were ALK+. There were no other significant differences in antigen expression between ALK+ and ALK cases. Discussion In the present study, we examined 29 cases of ALCL with the intent of better characterizing the FC immunophenotype of this lymphoma type, assessing the sensitivity of Downloaded from Am J Clin Pathol 2007;128:

5 Kesler et al / FLOW CYTOMETRY OF ANAPLASTIC LYMPHOMA CD8-PE CD4-FITC IgG1-PE CD56-PE IgG1-PE CD3-APC CD2-PE CD19-APC IgG1-FITC CD5-PerCP IgG2a-FITC Image 4 (Case 22) This anaplastic lymphoma kinase positive anaplastic large cell lymphoma (red) from a 7-year-old boy is notable for a light scatter that is somewhat diffuse but that concentrates approximately in the monocyte region. This neoplasm was CD30+/CD45+ but was negative for all T-cell antigens except for very dim expression of CD4 (compared with the G 1 FITC isotype control). CD13 was also positive (not shown). FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. Image 5 (Case 7) This anaplastic lymphoma kinase positive anaplastic large cell lymphoma (red) from a 19-year-old woman demonstrates high orthogonal light scatter with a diffuse pattern. The tumor cells are CD30+/CD45+/ CD2dim+/CD3dim+/CD4dim+/CD5 /CD7. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. FC analysis for its detection and comparing antigen expression in ALK+ and ALK cases. All ALCL cases were immunophenotypically aberrant with respect to T-cell antigen expression (CD2, CD3, CD4, CD5, CD7, and CD8). The number of aberrancies ranged from 1 to 5 per case (average, 3.9). All cases expressed at least 1 T-cell marker, with the majority of cases expressing 2 or more. CD4 was expressed most commonly (80%), followed by CD2 (72%), CD3 (40%), and CD5 and CD7 (32% each). T-cell antigens were often expressed at an abnormal intensity compared with normal T cells, with dim antigen expression predominating (Table 1). No case expressed CD8. The presence or absence of ALK expression divides ALCLs into 2 prognostically and clinicopathologically distinct groups. We were thus interested in determining whether immunophenotypic differences exist between ALK+ and ALK tumors. Overall, ALK+ and ALK ALCLs had similar immunophenotypic features. However, we found that ALK+ ALCL cases were more likely to express CD4 than ALK cases; in fact, all ALK+ cases were CD4+, although the majority of ALK cases were as well. Notably, however, no ALK cases expressed CD4 at the level seen in normal CD4+ T cells, whereas nearly half of the ALK+ cases showed normal CD4 intensity. We additionally found that all 5 CD13+ cases 318 Am J Clin Pathol 2007;128: Downloaded 318 from

6 Hematopathology / ORIGINAL ARTICLE CD56-PE CD5-FITC CD2-FITC assessed for ALK were positive, whereas both CD13 ALCLs were ALK. Thus, it seems that there may be immunophenotypic differences between ALK+ and ALK ALCLs, although IgG1-PE IgG2a-FITC Image 6 (Case 21) This anaplastic lymphoma kinase negative anaplastic large cell lymphoma (red) from a 51-year-old man shows high orthogonal light scatter similar to granulocytes. The tumor cells are CD30+/CD45. In addition, the lymphoma cells are CD2bright+/CD3 /CD4bright+/CD5bright+/CD7. FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein. the numbers of cases in this study are too few to draw definitive conclusions. CD30 has been reported to be positive by FC analysis in 60% to 100% of cases. 17,19,20 We were able to demonstrate CD30 in 23 (92%) of 25 cases. The remaining 2 cases were positive for CD30 by immunohistochemical analysis but negative by FC analysis, even though the same clone (Ber-H2, DAKO) was used for both analyses; there is no clear explanation for this phenomenon. In one case, the neoplastic cells showed a spectrum of size, and CD30 was expressed on only the large neoplastic cells by immunohistochemical analysis, whereas the population detected by FC analysis was in the medium-sized lymphocyte region (and accounted for 57% of events analyzed). Therefore, it is possible that the larger CD30+ cells were selectively lost in processing. In the other case, CD30 staining was weak and patchy by immunohistochemical analysis, but ALK was diffusely and strongly positive. In this case, the aberrant population identified by FC analysis was in the large lymphocyte region and accounted for only 3% of events. Expression of myeloid antigens has been previously reported in ALCL and raises the differential diagnosis of myeloid sarcoma. 9-12,18,20 In our series, aberrant expression of myeloid antigens was common in cases assessed for these markers. Most notably, CD13 expression was seen in 7 of 9 cases assessed. Six of the CD13+ cases were negative for CD3, and the seventh showed only partial dim CD3 expression. In addition, all CD13+ cases were CD4+. The fact that many tumors expressed myeloid antigens in the absence of CD3 serves to underscore the point that ALCL may easily be misdiagnosed as myeloid sarcoma, especially if a limited analysis is performed. CD4 expression in these tumors further complicates the differential diagnosis because many monocytic sarcomas are CD4+. Expression of CD15 in 2 of 4 cases in which it was assessed is surprising, given the infrequency of detection of CD15 by immunohistochemical analysis. 10,12,15,16,24 Compared with previous FC studies on ALCL, several of our results are similar. For example, Gorczyca et al 19 and Juco et al 20 also found that CD2 was commonly expressed in ALCL (71%-100% of cases; 72% in the current series), whereas CD3 Table 1 T-Cell Antigens in 25 Cases Positive by Flow Cytometric Analysis * CD2 CD3 CD5 CD7 CD4 Total positive 18 (72) 10 (40) 8 (32) 8 (32) 20 (80) Normal 9 (50) 1 (10) 4 (50) 2 (25) 8 (40) Dim 7 (39) 9 (90) 1 (13) 5 (63) 9 (45) Bright 2 (11) 0 (0) 3 (38) 1 (13) 3 (15) Total negative 7 (28) 15 (60) 17 (68) 17 (68) 5 (20) * Data are given as number (percentage). Downloaded from Am J Clin Pathol 2007;128:

7 Kesler et al / FLOW CYTOMETRY OF ANAPLASTIC LYMPHOMA Table 2 Comparison of ALK+ and ALK Cases for T-Cell Antigen Expression ALK+ (n = 13) ALK (n = 8) CD2 Normal intensity 4 5 Underexpressed 6 1 Overexpressed 0 1 Negative 3 1 CD3 Normal intensity 1 0 Underexpressed 4 4 Overexpressed 0 0 Negative 8 4 CD5 Normal intensity 3 1 Underexpressed 1 0 Overexpressed 1 1 Negative 8 6 CD7 Normal intensity 0 1 Underexpressed 4 1 Overexpressed 1 0 Negative 8 6 CD4 Normal intensity 6 0 Underexpressed 5 4 Overexpressed 2 1 Negative 0 3 ALK, anaplastic lymphoma kinase. and CD7 were less frequently positive (25%-32%; 32%-40% in the current series). CD30 expression has been described in 60% to 100% of cases, compared with 92% of our cases. 17,19,20 Likewise, CD4 expression seems to be consistently more common than CD8 expression (50%-63%; 80% in the current series). 19,20 However, a number of our findings are different from those of previous investigators. We found a higher percentage of cases expressing CD13 (78% vs 47% 20 ) and CD56 (60% vs 0% 19 ). In addition, although previous studies described ALCL cases as uniformly CD45+ with bright to moderate intensity, 8% of cases in the current series were CD45. Furthermore, 22% of the CD45+ cases underexpressed CD45 relative to normal lymphocytes. Finally, none of the cases in our series expressed CD8; 80% of cases were CD4+, and the remaining 20% were double negative (CD4 /CD8 ). This is in contrast with the findings of Gorczyca et al 19 (25% of cases CD8+, 25% double negative) and Juco et al 20 (16% of cases CD8+, 5% double positive, and 21% double negative). Previous studies did not address any immunophenotypic differences between ALK+ and ALK cases. Detection of lymphomas by FC analysis is dependent on a number of factors, including adequate sampling of the involved node or tissue, specimen viability, and the ability of tumor cells to survive processing. False-negative FC results are especially common in large cell lymphomas; it is estimated that up to 25% of large cell lymphomas are not detected by FC analysis. 21 The reduced sensitivity of FC analysis in large cell lymphomas is presumably due to the necrosis and apoptosis commonly associated with such tumors and to sampling issues and cellular fragility resulting in selective loss of tumor cells during processing. 17,21,25 In our series, ALCL was identified by FC analysis in 25 (86%) of 29 cases. In 1 of the cases with negative FC results, malignant cells were not morphologically evident on the cytocentrifuged preparation, suggesting sampling error. In another case, the large neoplastic cells were few, and it is possible that they were lost in processing. The remaining 2 cases seem to be false-negative cases. Gating and data analysis strategies also have a role in the ability to detect large malignant cells. A common approach for the assessment of possible lymphoma in FC specimens is to gate on and analyze only events in the low orthogonal scatter regions of the forward scatter/orthogonal scatter or CD45/orthogonal scatter histograms. This approach would have failed to identify the majority of ALCL cases in this series. In 16 (64%) of 25 cases, the ALCL cells had high orthogonal scatter, similar to monocytes or granulocytes, with variable forward scatter. Only 9 cases (36%) had low orthogonal scatter, similar to lymphocytes or blasts. Cases with high orthogonal scatter also often failed to form tight light scatter clusters. If gating had been performed based solely on scatter characteristics, the tumor cells would likely have been missed, especially when tumor cells were few. Even examination of all CD45+ events would have resulted in some false-negatives because the tumor cells in 2 of the cases (8%) lacked CD45. Our data were ungated in all cases, allowing for the identification of clusters based not only on CD45 and light scatter properties, but also on any fluorescence or fluorescence/light scatter plots. For example, the presence of a discrete CD30+ cluster or a CD4+ population that lacked CD3 might alert the operator to the existence of an abnormal population. This population could then be specifically analyzed with respect to scatter properties and other antigens in those analysis tubes. With respect to the differential diagnosis of CD30+ populations by FC analysis, several points are worth highlighting. First, although CD30+ T cells may be seen in the paracortex of reactive lymph nodes by immunohistochemical analysis, in our experience, discrete populations of CD30+ T cells are not seen in reactive tissues by FC analysis, 23 although one would assume they would be detectable in small numbers. If such a population was detected, the lack of other immunophenotypic aberrancy would presumably allow distinction from ALCL. Second, it is clear that expression of CD30 is not specific for ALCL among peripheral T-cell lymphomas, and the appropriate morphologic features need to be present to confirm this diagnosis. Third, B-lineage NHLs may express CD30. 26,27 The expression of B-lineage markers in most cases will help easily discriminate these from ALCL. However, rare immunophenotypically null, CD30+ lymphomas of B-lineage, 320 Am J Clin Pathol 2007;128: Downloaded 320 from

8 Hematopathology / ORIGINAL ARTICLE in particular human herpesvirus 8 associated primary effusion lymphoma, may aberrantly express T cell associated antigens (eg, CD2, CD4, and CD7). 28,29 In the 1 case in the present study that manifested in pleural fluid, primary effusion lymphoma could be confidently excluded based on expression of CD5, CD56, and CD13. In addition, the patient was an 8-year-old girl. Thus, correlation with clinical, morphologic, and other immunophenotypic features is essential to rule out this differential diagnosis. Finally, classical Hodgkin lymphoma is a theoretical differential diagnostic issue when CD30+ cells are found by FC analysis. However, despite the recent report of detection of Reed- Sternberg cells identified by FC analysis using a specialized approach, 30 we have been consistently unable to identify a discrete CD30+ neoplastic population by FC analysis in cases of classical Hodgkin lymphoma. Although FC analysis is able to detect ALCL in tissue biopsy specimens in most cases, it is likely that it will remain essentially an incidental diagnostic modality in the workup of this type of lymphoma. The main advantage of FC analysis over immunohistochemical analysis at the time of initial diagnostic workup is speed, whereby a diagnosis may be available well in advance of that obtainable through routine immunohistochemical workup, even before the review of H&E-stained slides. In our experience, the diagnosis was suspected in many cases from review of the processed cytocentrifuged slide at the time of initial triage. However, the routine antibody panel for tissue specimens during the period of this study included CD30 in a tube with CD3, CD2, and CD45, so that even unsuspected cases were on occasion identified during data analysis. Diagnosis by FC analysis thus preceded review of H&E-stained slides in most of our cases, allowing a streamlining of the subsequent immunohistochemical workup (eg, restrict it to assessment of ALK expression). FC analysis may also be helpful when material for immunohistochemical analysis is not available, such as in the case of fine-needle aspiration. Conclusions The majority of ALCLs may be detected by FC analysis, provided that appropriate gating and data analysis techniques are used; reliance on light scatter and/or CD45-based gating strategies may result in false-negative FC results. All ALCL cases are immunophenotypically aberrant, but all in the present series expressed at least 1 T-cell antigen, most commonly CD4. With the exception of differences in CD4 expression and, possibly, myeloid antigen expression, ALK+ and ALK tumors seem to have largely similar immunophenotypic characteristics. Finally, expression of myeloid antigens, particularly CD13, is common in ALCL, and this, together with frequent lack of common T-cell antigens, may lead to the misdiagnosis of these tumors as myeloid sarcomas if only limited analyses are performed. From the Department of Pathology, University of Texas Southwestern Medical Center, Dallas. Address reprint requests to Dr Kroft: Dept of Pathology, Medical College of Wisconsin, 8701 Watertown Plank Rd, Milwaukee, WI References 1. Rimokh R, Magaud JP, Berger F, et al. A translocation involving a specific breakpoint (q35) on chromosome 5 is characteristic of anaplastic large cell lymphoma ( Ki-1 lymphoma ). Br J Haematol. 1989;71: Morris SW, Kirstein MN, Valentine MB, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in non- Hodgkin s lymphoma. Science. 1994;263: Falini B, Pileri S, Zinzani PL, et al. ALK+ lymphoma: clinicopathological findings and outcome. Blood. 1999;93: Shiota M, Nakamura S, Ichinohasama R, et al. Anaplastic large cell lymphomas expressing the novel chimeric protein p80npm/alk: a distinct clinicopathologic entity. Blood. 1995;86: ten Berge RL, de Bruin PC, Oudejans JJ, et al. ALK-negative anaplastic large-cell lymphoma demonstrates similar poor prognosis to peripheral T-cell lymphoma, unspecified. Histopathology. 2003;43: Gascoyne RD, Aoun P, Wu D, et al. Prognostic significance of anaplastic lymphoma kinase (ALK) protein expression in adults with anaplastic large cell lymphoma. Blood. 1999;93: Foss HD, Anagnostopoulos I, Araujo I, et al. Anaplastic largecell lymphomas of T-cell and null-cell phenotype express cytotoxic molecules. Blood. 1996;88: Bonzheim I, Geissinger E, Roth S, et al. Anaplastic large cell lymphomas lack the expression of T-cell receptor molecules or molecules of proximal T-cell receptor signaling. Blood. 2004;104: Popnikolov NK, Payne DA, Hudnall SD, et al. CD13-positive anaplastic large cell lymphoma of T-cell origin: a diagnostic and histogenetic problem. Arch Pathol Lab Med. 2000;124: Gorczyca W, Tsang P, Liu Z, et al. CD30-positive T-cell lymphomas co-expressing CD15: an immunohistochemical analysis. Int J Oncol. 2003;22: Clavio M, Rossi E, Truini M, et al. Anaplastic large cell lymphoma: a clinicopathologic study of 53 patients. Leuk Lymphoma. 1996;22: Perkins PL, Ross CW, Schnitzer B. CD30-positive, anaplastic large-cell lymphomas that express CD15 but lack CD45: a possible diagnostic pitfall. Arch Pathol Lab Med. 1992;116: Martinez B, Gonzalo I, Robledo M, et al. Large-cell anaplastic lymphomas: a genetic and immunophenotypic study [in Spanish]. Sangre (Barc). 1995;40: Ohshima K, Kikuchi M, Masuda Y, et al. Genotypic and immunophenotypic analysis of anaplastic large cell lymphoma (Ki-1 lymphoma). Pathol Res Pract. 1990;186: Downloaded from Am J Clin Pathol 2007;128:

9 Kesler et al / FLOW CYTOMETRY OF ANAPLASTIC LYMPHOMA 15. Suzuki R, Kagami Y, Takeuchi K, et al. Prognostic significance of CD56 expression for ALK-positive and ALK- negative anaplastic large-cell lymphoma of T/null cell phenotype. Blood. 2000;96: Felgar RE, Salhany KE, Macon WR, et al. The expression of TIA-1+ cytolytic-type granules and other cytolytic lymphocyteassociated markers in CD30+ anaplastic large cell lymphomas (ALCL): correlation with morphology, immunophenotype, ultrastructure, and clinical features. Hum Pathol. 1999;30: Dunphy CH. Contribution of flow cytometric immunophenotyping to the evaluation of tissues with suspected lymphoma? Cytometry. 2000;42: Dunphy CH, Gardner LJ, Manes JL, et al. CD30+ anaplastic large-cell lymphoma with aberrant expression of CD13: case report and review of the literature. J Clin Lab Anal. 2000;14: Gorczyca W, Weisberger J, Liu Z, et al. An approach to diagnosis of T-cell lymphoproliferative disorders by flow cytometry. Cytometry. 2002;50: Juco J, Holden JT, Mann KP, et al. Immunophenotypic analysis of anaplastic large cell lymphoma by flow cytometry. Am J Clin Pathol. 2003;119: Bertram HC, Check IJ, Milano MA. Immunophenotyping large B-cell lymphomas: flow cytometric pitfalls and pathologic correlation. Am J Clin Pathol. 2001;116: Xu Y, McKenna RW, Kroft SH. Comparison of multiparameter flow cytometry with cluster analysis and immunohistochemistry for the detection of CD10 in diffuse large B-cell lymphomas. Mod Pathol. 2002;15: Jamal S, Picker LJ, Aquino DB, et al. Immunophenotypic analysis of peripheral T-cell neoplasms: a multiparameter flow cytometric approach. Am J Clin Pathol. 2001;116: Tilly H, Gaulard P, Lepage E, et al. Primary anaplastic largecell lymphoma in adults: clinical presentation, immunophenotype, and outcome. Blood. 1997;90: Jennings CD, Foon KA. Recent advances in flow cytometry: application to the diagnosis of hematologic malignancy. Blood. 1997;90: Higgins JP, Warnke RA. CD30 expression is common in mediastinal large B-cell lymphoma. Am J Clin Pathol. 1999;112: Lai R, Medeiros LJ, Dabbagh L, et al. Sinusoidal CD30- positive large B-cell lymphoma: a morphologic mimic of anaplastic large cell lymphoma. Mod Pathol. 2000;13: Shamah CJ, Pretell JO, Kroft SH. Pericardial effusion in a 29-year-old man. Lab Med. 2007;38: Wilson KS, McKenna RW, Kroft SH, et al. Primary effusion lymphomas exhibit complex and recurrent cytogenetic abnormalities. Br J Haematol. 2002;116: Fromm JR, Kussick SJ, Wood BL. Identification and purification of classical Hodgkin cells from lymph nodes by flow cytometry and flow cytometric cell sorting. Am J Clin Pathol. 2006;126: Am J Clin Pathol 2007;128: Downloaded 322 from