- Email: [email protected]
T-cell lymphoma, nasal type
Human Pathology (2014) 45, 2255–2262
www.elsevier.com/locate/humpath
Original contribution
Identification of immunophenotypic subtypes with different prognoses in extranodal natural killer/T-cell lymphoma, nasal type☆,☆☆ Jian-Bo Yu PhD a,b,1 , Zhuo Zuo PhD a,1 , Wen-Yan Zhang PhD a , Qun-Pei Yang PhD a , Ying-Chun Zhang PhD a , Yuan Tang PhD a , Sha Zhao PhD a , Xian-Ming Mo PhD a , Wei-Ping Liu MD a,⁎ a
Department of Pathology and Laboratory of Stem Cell Biology, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, 610041, China b Department of Pathology, Key Laboratory of Cancer Prevention and Treatment of Heilongjiang Province, Mudanjiang Medical University, Mudanjiang, 157011, China Received 14 March 2014; revised 21 April 2014; accepted 24 April 2014
Keywords: CD16; CD56; Lymphoma; Natural killer cells; T lymphocytes
Summary To analyze the differentiation characteristics of extranodal natural killer/T-cell lymphoma, nasal type, one nude mouse model, cell lines SNK6 and SNT8, and 16 fresh human samples were analyzed by flow cytometry immunophenotyping and immunohistochemistry staining; and 115 archived cases were used for phenotypic detection and prognostic analysis. We found that CD25 was expressed by most tumor cells in all samples, and CD56+CD25+ cells were the predominant population in the mouse model, the 2 cell lines, and 10 of the 16 fresh tumor samples; in the other 6 fresh tumor samples, the predominant cell population was of the CD16+CD25+ phenotype, and only a minor population showed the CD56+CD25+ phenotype. The phenotype detected by immunohistochemistry staining generally was consistent with the phenotype found by flow cytometry immunophenotyping. According to the expression of CD56 and CD16, 115 cases could be classified into 3 phenotypic subtypes: CD56−CD16−, CD56+CD16−, and CD56dim/−CD16+. Patients with tumors of the CD56dim/−CD16+ phenotype had a poorer prognosis than patients with tumors of the other phenotypes. Differentiation of extranodal natural killer/T-cell lymphoma, nasal type apparently resembles the normal natural killer cell developmental pattern, and these tumors can be classified into 3 phenotypic subtypes of different aggressiveness. Expression of CD56dim/−CD16+ implies a poorer prognosis. © 2014 Elsevier Inc. All rights reserved.
☆ Competing interests: The authors have declared no potential conflicts of interest. ☆☆ Funding/Support: This work was supported by The National Natural Science Foundation of China (Beijing) (no. 81272626 and 81271628) and the Innovation Team Project of Tumor Prevention and Treatment in Heilongjiang University, Health Department Foundation of Heilongjiang Province (no. 2009-412). ⁎ Corresponding author. Department of Pathology, West China Hospital of Sichuan University, No. 37, Guoxue Alley, Chengdu 610041 Sichuan, China. E-mail address: [email protected] (W. -P. Liu). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.humpath.2014.04.024 0046-8177/© 2014 Elsevier Inc. All rights reserved.
1. Introduction Human extranodal natural killer (NK)/T-cell lymphoma, nasal type (ENKTL), is a distinct clinicopathological entity associated with a highly aggressive clinical course and thus
2256
JB. Yu et al. ative support for investigations of the differentiation patterns of ENKTL. In human bone marrow and secondary lymphoid tissue (SLT), CD34+ hematopoietic stem cells and hematopoietic progenitor cells can give rise to cytolytic NK cells after stimulation with either interleukin (IL)-2 or IL-15 [10,11]. Freud et al [12] used the combination of CD34, CD117, and CD94 antigens to discriminate 4 functionally distinct stages of human CD56bright NK-cell development within SLT. Some reports demonstrated that both CD56 bright CD16 − and CD56dim/−CD16bright cells are mature NK cells, whereas CD56brightCD16− NK cells could be the direct precursors of CD56dim/−CD16bright NK cells [13-17]. Some authors believe that CD56brightCD16− and CD56dim/−CD16bright NK cells are derived from SLT and bone marrow, respectively [14-17]. Whether the CD56brightCD16− and CD56dim/−CD16bright phenotypes occur in ENKTL remains uncertain. In this study, based on reports describing normal NKcell development, differentiation patterns of ENKTL were investigated. The samples included a nude mouse model, 2 ENKTL cell lines, 16 fresh tissue samples, and 115 paraffin-embedded block samples from human ENKTL. Immunophenotypic analysis by both flow cytometry immunophenotyping (FCI) and immunohistochemistry (IHC) staining demonstrated that tumor-cell populations in ENKTL represent a continuum of normal stages of
a poor prognosis [1]. It is rare in Western countries but more common in Asia, Mexico, and Central and South America [1,2]. The clinical features are more aggressive, and the prognosis is poorer in extranasal than in nasal tumors [3]. Most of the tumors exhibit NK-cell phenotypes with expression of CD56 and cytoplasmic CD3 but do not express surface CD3; and T-cell receptor (TCR) genes remain in the germline configuration [4]. However, some cases have a clonal T-cell phenotype and genotype [5,6]. The differentiation patterns of lymphatic and hematopoietic leukemia have been well described, but the differentiation patterns of lymphomas, including ENKTL, are less clear. A series of articles reported that hematopoietic stem cells and committed progenitor cells could transform into tumorinitiating cells and differentiate along the original lineage [7]. B-cell acute lymphoid leukemia is phenotypically reminiscent of the normal stages of B-cell differentiation, and a scheme that assigns stages of human B-cell acute lymphoid leukemia similar to those of normal B cells has been proposed [8]. This developmental pattern suggests that leukemia represents normal hematopoietic development gone slightly awry [9]. Some studies have elucidated the process of normal NK-cell development [10-17]. These studies provide comparTable 1
Clinicopathological features and immunophenotype of 16 human tumor samples, nude mouse model, and cell lines of ENKTL
Sample no.
Human tumor 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Mouse model SNK6 [18] SNT8 [18] NK-YS [19] HANK-1 [20]
Sex/ age
Primary lesion
F/42 M/42 M/39 F/35 M/59 M/18 F/53 M/33 M/33 M/19 M/63 F/51 F/25 F/62 M/37 M/42 M/48 M/62 F/48 F/19 F/46
Nasal Nasal Nasal Skin Colon Palate Palate Nasal Palate Nasal Nasal Colon Palate Nasal Pharynx Skin Nasal Nasal Nasal Nasal Retroperitoneal
IHC
FCI
CD3ε
CD20
CD56
GB
+ + + + + + + + + + + + + + + + + + + + +
− − − − − − − − − − − − − − − − − − − − NA
+ + − + + + + + − − − + − − + + + + + + +
+ + + + + + + + + + + + + + + + + NA NA NA +
CD56+CD16− (%) 67 87 22 67 39 88 66 93 6 5 15 90 12 17 64 89 90 86 76 + +
CD56dim/−CD16+ (%) NA NA 54 0 0 NA NA 0 90 91 70 0 82 33 0 0 0 0 0 0 0
TCR rearr. − − − − + NA NA − + − − − − − NA − − − + − −
NOTE. All specimens were positive for EBV. The percentage of CD56+CD16− and CD56dim/-CD16+ cells was calculated under CD25+ gating; SNK6, SNT8, NK-YS, and HANK-1 are ENKTL cell lines. Only flow cytometry analysis was done in the present study; the other information is from the literature. All of the information on NK-YS and HANK-1 is from the literature. Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; IHC, immunohistochemistry; FCI, flow cytometry immunophenotyping; NA, not available; GB, granzyme B; rearr., rearrangement.
Phenotypic subtypes in ENKTL human NK-cell differentiation; and CD56 − CD16 − , CD56+ CD16− , and CD56 dim/− CD16+ phenotypic subtypes were identified. The CD56 dim/− CD16 + subtype was associated with a poorer prognosis.
2. Materials and methods 2.1. Patients Sixteen fresh samples of human ENKTL were collected from West China Hospital of Sichuan University for FCI analysis (Table 1). Before collection of these samples, none of the patients had received any antitumor treatment. Informed consent was obtained from all enrolled patients, and the procedure for tissue collection was approved by the Human Research Ethical Committees of West China Hospital, Sichuan University. In addition, archived samples harvested from 115 patients with ENKTL between January 1993 and December 2010 were obtained from the Pathology Department of West China Hospital and used for clinicopathological review, IHC staining, and prognostic analysis. The diagnosis of all cases was made by 2 experienced hematopathologists using the updated World Health Organization classification of tumors of the hematopoietic and lymphoid tissues [1].
2.2. Nude mouse tumor model and tumor cell lines A BALB/c nude mouse model of human ENKTL was established using secondary gastric NK/T-cell lymphoma and passaged as described previously [21]. Two ENKTL cell lines, SNK6 and SNT8, were kindly provided by Prof Nagata [18]. The SNK6 cells expressed the phenotype CD3−CD4− CD8−CD56+, and TCRγ genes were not rearranged; SNT8 expressed the phenotype CD3 + CD4 − CD8 − CD56 + and exhibited TCRγ rearrangement. Both SNK6 and SNT8 were cultured in RPMI 1640 (Gibco, Grand Island, NY) medium supplemented with antibiotics, 10% heat-inactivated human serum, and recombinant human IL-2 600 U/mL and were incubated in a humidified atmosphere of 5% carbon dioxide at 37°C.
2.3. Cell preparation Neoplastic tissues from human ENKTL or from the BALB/c nude mice were cut into thin slices, suspended in RPMI 1640 medium, and digested using collagenase 3 (Worthington Biochemical, Lakewood, NJ). The mixture was incubated at 37°C for as long as 3 hours to allow complete digestion. Cells were filtered through 40-μm nylon mesh and washed twice with phosphate-buffered saline (PBS) supplemented with 2% fetal calf serum and 0.1% sodium azide and then stained for FCI analysis.
2257
2.4. Antibody staining for FCI Suspensions of mononuclear cells prepared by tissue digestion were incubated with antibodies for 20 minutes at 4°C, washed with PBS, resuspended in PBS, and analyzed by FCI. Fluorochrome-conjugated monoclonal antibodies against the following antigens were purchased from PharMingen (Becton Dickinson, San Diego, CA): CD3 (UCHT1), CD4 (RPA-T4), CD7 (M-T701), CD8 (SK1), CD14 (M5E2), CD16 (3G8), CD20 (2H7), CD25 (M-A251), CD31 (WM59), CD34 (581), CD38 (HIT2), CD44 (G4426), CD45 (H130), CD56 (B159), CD94 (HP-3D9), CD117 (YB5.B8), CD122 (Mik-β2), HLA-DR (TU36), integrin β7 (FIB504), and isotype control immunoglobulin G (R3-34). Dead cells were distinguished by staining with 7-amino-actinomycin D. All fresh samples and cell lines were gated according to their light scatter characteristics. Data were analyzed using FACS/Calibour CELLQuest, Becton Dickinson (San Jose, CA).
2.5. IHC staining Antibodies used for IHC included CD3ε (polyclonal; Dako, Glostrup, Denmark), CD16 (275003; R&D, Minneapolis, MN), CD20 (L26; Dako), CD25 (IL2R.1; Zymed, Grand Island, NY), CD34 (QBEnd/10; NeoMarker, Fremont, CA), CD45RO (UCHL1; Dako), CD56 (123, C3; Zymed), granzyme B (GZB01; NeoMarkers), and CD117 (2E4; Zymed). Sections (4 μm) were cut and deparaffinized in xylene and then hydrated in a graded series of ethanol. The slides were treated by pressure cooking in citric acid buffer (10 mmol/L, pH 7.4) for 3 minutes before staining. Bound antibodies were visualized by the EnVision (Dako) method using 3,3′-diaminobenzidine as a substrate.
2.6. In situ hybridization and polymerase chain reaction The in situ hybridization (ISH) was carried out with a fluorescein-labeled oligonucleotide probe complementary to 2 Epstein-Barr virus (EBV)–Epstein-Barr virus encoded small RNAs 1 and 2 (EBER1/2) (Y520001; Dako). Rabbit antifluorescein isothiocyanate (FITC) antibody conjugated with alkaline phosphatase (Dako) was used to detect the probe. Genomic DNA from paraffin-embedded tissue samples was extracted by phenol-chloroform procedures. Both universal TCRγ (TVG/TJX) primers and BIOMED-2 polymerase chain reaction (PCR) multiplex tubes (TCRβ + TCRγ) were used for TCR rearrangement analysis by PCR. These procedures were performed as we previously reported [22].
2.7. Survival and statistical analysis Follow-up data were available for all 115 patients. The overall survival (OS) time was plotted using the Kaplan-Meier
2258 procedure (SPSS 13.0; SPSS, Chicago, IL). Differences in survival times between patient groups were calculated using the log-rank test. P b .05 was considered statistically significant.
3. Results 3.1. Differentiation characteristic detected by FCI in ENKTL For discovering the differentiation characteristics of ENKTL, a nude mouse tumor model, 2 cell lines, and 16 samples of fresh human ENKTL were used for immunophenotypic analysis by FCI using a series of antibodies. We found that CD25 was expressed by a majority of the tumor cells in all samples (Fig. 1A), and some of these cells coexpressed CD56, CD117, and CD122. We also observed some CD117+ cells also expressing CD56. In the mouse model and 2 cell lines, CD25+ CD56+ cells were the predominant population (Fig. 1B), and CD25+CD56− cells were a minor population. In the SNK6 and SNT8 tumor cell lines, the phenotype features of the lines changed during culture: after 24 hours, CD56+ tumor cells represented 30% of the total, but this figure increased to more than 76% by 72 hours (Fig. 1C and D), and CD34+ cells were observed consistently in small numbers (Fig. 1E). However, when we analyzed 16 fresh human tumor samples, a predominance of CD56+CD25+ cells was observed in only 10, and there was no CD16 expression in these cases (Fig. 1F). In the other 6 samples (see Table 1), a lower frequency of CD56
JB. Yu et al. expression was noticed, and further analysis revealed that the phenotype of the predominant cells was CD16+ (Fig. 1G). In the CD16+ samples, a minor CD56+CD16− cell population could be observed, but no CD20+ or CD14+ cells were present. The CD117 (c-kit) protein typically is expressed by precursor and progenitor cells, and CD122 is the IL-2/IL-15 receptor β chain and a marker for immature NK cells [10]. Both were detected in minor cell populations in all samples (Fig. 1E and H). Expression of CD94 was low, and a small percentage of CD3+ tumor cells was observed in cases 5 and 9. CD8+ tumor cells were noted in cases 5, 9, 10, 11, and 14. No CD4 expression was observed in any cells.
3.2. Immunophenotype characteristics detected by IHC in ENKTL We also assessed phenotypic characteristics using CD56 and CD16 markers detected by IHC (Fig. 2). Expression of CD56 and CD16 generally coincided with the phenotypes found by FCI. However, when the case with no more than 22% of CD56dim tumor cells detected by FCI, only weak or no CD56 expression was observed by IHC (see Table 1). This might be the reason that a low density of CD56 antigen on tumor cells is difficult to detect using the IHC method. Based on the expression characteristics of the CD56 and CD16 markers, the tumors could be classified into 3 phenotypic subtypes. The first, no matter weak or no CD56 expression (CD56dim/− ), the CD16+ cases were classified as the CD56dim/−CD16+ phenotype, similar to normal NK cells [13-15]. The second was CD56+ cases without CD16 expression, which was classified as the
Fig. 1 Immunophenotypic features of ENKTL detected by FCI. Dead cells were excluded by 7-amino-actinomycin D staining. Dot plots were gated based on light scatter characteristics and isotype controls. A, Most tumor cells in the nude mouse model were CD25+ and CD16−. B, CD56+CD25+ tumor cells were the predominant population in the nude mouse model. C, After 24 hours in culture, a minority of proliferating CD56+ cells were observed in SNK6. D, A majority of cells expressing CD56 were observed after 72 hours in culture. E, A minimum of CD34+ and CD122+ cells were observed in the 2 cell lines. F, No CD16+ cells were observed in 10 of 16 fresh tumor samples. G, Of 16 fresh tumor samples, 6 were CD16+, and a fraction of the cells expressed the CD56+CD16− phenotype. H, CD117+ tumor cells were observed in all samples, some of these cells were CD25−.
Phenotypic subtypes in ENKTL
2259
Fig. 2 Immunophenotypic features detected by IHC. A to C, Tumor cells with CD56+CD16− phenotype. A, Strong expression of CD56. B, Cells are negative for CD16. C, Granzyme B was moderately expressed. D to F, CD56dim/−CD16+ cases. D, No expression of CD56. E, Expression of CD16. F, Strong expression of granzyme B. Original magnification, × 200.
CD56+CD16− phenotype. The third was CD56− cases without CD16 expression, which was classified as a CD56−CD16− phenotype. The CD25, CD34, and CD117 markers also were sought by IHC, but no positive tumor cells were observed. Expression of granzyme B was observed in all 3 phenotypes with no significant difference among the phenotypes. In addition, EBER1/2 staining by ISH was positive in all cases. Case 5 (CD56+CD16− phenotype) and case 9 (CD56dim/− CD16+ phenotype) contained TCRγ gene rearrangements. No other cells expressed this feature.
3.3. Clinicopathological characteristics and prognostic analysis of 3 subtypes of ENKTL Archived data of 115 ENKTL cases were reviewed (Table 2). The patients included 87 male and 28 female subjects with a median age of 40 years (range, 6-84 years). Forty-one patients (35.7%) had B symptoms. Ninety-one patients (79.1%) had tumors in the nasal/upper aerodigestive tract, and 24 (20.9%) presented with extranasal sites of involvement as follows: small intestine (n = 5), colon (n = 8), skin (n = 8), and lymph nodes (n = 3). Sixteen patients (13.9%) were in stage III/IV. Six patients (5.2%) had tumors at more than one site. All cases showed coagulative necrosis to various degrees. Angiocentric infiltration was observed in 15 cases (13%). The tumors showed predominantly medium-sized cells (n = 105 [91.3%]). In 10 cases (8.7%), the tumor cells were predominantly large. Six patients were treated with radiation therapy, 24 patients with chemotherapy, 15 patients with operation and chemotherapy, and 4 patients with antiinflammatory therapy; and 45 patients received combined radiation
and chemotherapy. The most commonly used chemotherapy regimen was cyclophosphamide, doxorubicin, vincristine, and prednisolone. Treatment information was not available for 21 patients. Follow-up data were available for all 115 patients (range, 0.08-5.0 years). With a median follow-up of 0.92 years, 59 patients (51.3%) had died. The median OS time was 1.31 years, and the OS was 48.7%. The results of IHC are summarized in Table 2. All cases were CD20−. CD3ε was positive in 100 of 102 cases, granzyme B was positive in 110 of 111 cases, CD56 was positive in 90 of 115 cases, and CD16 was positive in 19 of 115 cases. EBV-positive cells were detected by ISH in all 115 cases. According to the expression of CD56 and CD16, 115 cases were classified into 3 phenotypic groups, 6 CD56− CD16−, 90 CD56+CD16−, and 19 CD56dim/−CD16+. Follow-up data were used for prognosis analysis based on phenotypic subtypes, B symptom, lesions, lesion sites, and stages (Table 3, Fig. 3). The OS for the CD56+CD16− group was 57.8% compared with 33.3% for the CD56−CD16− group and 10.5% for the CD56dim/−CD16+ group (P b .0001). The patients with B symptoms had a poorer prognosis, although the difference was not statistically significant (P = .198). Stage III/IV disease, more than one lesion, and extranasal lesions correlated significantly with a poorer prognosis (P = .002, P = .0003, and P b .0001, respectively).
4. Discussion The ENKTL is the most common non–B-cell extranodal neoplasm seen in China, representing approximately 17% of all non-Hodgkin lymphomas [2]. This study was based on
2260 Table 2
JB. Yu et al. Clinicopathological data of 115 cases of ENKTL CD56−CD16−, n (%)
CD56+CD16−, n (%)
CD56dim/−CD16+, n (%)
115
6 (5.2)
90 (78.3)
19 (16.5)
87 (75.7) 28 (24.3)
5 (83.3) 1 (16.7)
67 (74.4) 23 (25.6)
15 (78.9) 4 (21.1)
40 6-84
38 25-60
43 15-84
36 6-69
41 (35.7) 74 (64.3)
1 (16.7) 5 (83.3)
31 (34.4) 59 (65.6)
9 (47.4) 10 (52.6)
10 (8.7) 105 (91.3)
1 (16.7) 5 (83.3)
8 (8.9) 82 (91.1)
1 (5.3) 18 (94.7)
15 (13) 100 (87)
1 (16.7) 5 (83.3)
10 (11.1) 80 (88.9)
4 (21.1) 15 (78.9)
100/102 (98.0) 110/111 (99.1) 0 115 (100)
4/4 (100) 6 (100) 0 6 (100)
79/81 (97.5) 86/86 (100) 0 90 (100)
17/17 (100) 18/19 (94.7) 0 19 (100)
99 (86.1) 16 (13.9)
5 (83.3) 1 (16.7)
81 (90) 9 (10)
13 (68.4) 6 (31.6)
91 (79.1) 24 (20.9)
6 (100) 0
76 (84.4) 14 (15.6)
9 (47.4) 10 (52.6)
109 (94.8) 6 (5.2%)
5 (83.3) 1 (16.7)
88 (97.8) 2 (2.2)
16 (84.2) 3 (15.8)
59 (51.3) 56 (48.7)
4 (66.7) 2 (33.3)
38 (42.2) 52 (57.8)
17 (89.5) 2 (10.5)
Total, n (%) Total no. Sex (%) Male Female Age (y) Median Range B symptoms (%) Yes No Cell size (%) Large Medium Angiocentric infiltration (%) Yes No IHC and EBV (%) CD3ε Granzyme B CD20 EBV Stage (%) I/II III/IV Lesion site (%) Nasal Extranasal Lesion size (cm) (%) 1 ≥2 Last known condition (%) Dead Alive
Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; IHC, immunohistochemistry; EBV, Epstein-Barr virus.
recent advances in delineating normal NK-cell development [10-17]. The expression of CD117 and CD34 marks precursor cells; CD56 expression in NK-cell development shows a dynamic process, which increases gradually from the immature stage to the CD56brightCD16− mature stage and then gradually wanes to CD56dim/−CD16+ [11,12]. In the present study, we observed minor cell populations with the CD117+ and CD34+ phenotypes in ENKTL, and CD56 expression gradually increased, as the duration of culture of SNK6 and SNT8 cells was extended. A minor percentage of the CD56+ tumor cell population was found in the CD56dim/− CD16+ phenotypic cases. These facts support the conclusion that the development process of the tumor represents a continuum of normal NK cells, even in the cases accompanied by TCRγ gene rearrangement; and the terminal point of tumor cell differentiation in different phenotypic subtypes may cease at different points of the developmental process from immature to mature cells. Craig and Foon [23] regarded flow cytometry as a valuable tool for identifying phenotypes and lineages and for the
Table 3
Prognostic analysis of 115 cases of ENKTL
B symptoms Yes No Lesions Nasal Extranasal Lesion size (cm) 1 ≥2 Stage I/II III/IV Phenotypic subtype CD56−CD16− CD56+CD16− CD56dim/−CD16+
n (%)
OS, n (%)
41 74
17 (41.5) 39 (52.7)
91 24
54 (59.3) 2 (8.3)
109 6
56 (51.4) 0
99 16
53 (53.5) 3 (18.8)
6 90 19
2 (33.3) 52 (57.8) 2 (10.5)
P .198
b.0001
.0003
.002
b.0001
Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; OS, overall survival.
Phenotypic subtypes in ENKTL
2261
Fig. 3 Prognosis analysis. OS analysis was performed on 115 cases with complete follow-up data. A, B symptoms were associated with shorter OS but without statistical significance. B, Patients whose tumors had the CD56dim/−CD16+ phenotype had poorer OS, whereas the CD56+CD16− phenotype was associated with a better OS. The results of prognosis analysis showed that stage III/IV (C), more than one lesion (D), and extranasal tumors (E) were significantly associated with a poorer prognosis.
diagnosis and evaluation of hematologic neoplasms. The CD25 protein is the IL-2 receptor α chain. Its ligand, IL-2, is an important cytokine for normal NK-cell growth [10-12]. In the present study, CD25 expression detected by FCI demonstrated that IL-2 is a key cytokine for ENKTL and the cell lines arising from it [18-20]. Furthermore, the immunophenotype detected by FCI demonstrated that not all CD56− cases were authentic T-cell–lineage tumors [24]. When CD56dim tumor cells were detected by FCI, the CD56 marker was difficult to detect by IHC, and only 1 of 6 CD56− cases (case 9) exhibited TCRγ rearrangement. Our finding indicates that CD56− cases in ENKTL might mimic normal NK cells with CD56dim/−CD16+ and CD56−/dimCD16− differentiation. Therefore, according to the expression of CD56 and CD16 detected by IHC, we suggest that when CD56− cases are accompanied by CD16 expression, the differentiation pathway of the tumor cells may pass through a CD56+CD16− stage on their way to the terminal CD56dim/− CD16+ stage. When the CD56 marker was expressed, the differentiation of the tumor cells may cease at the CD56+CD16− mature stage; when CD56− cases did not manifest CD16 expression, differentiation may have ceased at the early CD56+CD16− mature stage or a later immature stage. Tumors not expressing CD56 or CD16 have been recognized for more than a decade [4,5,25-28]. In the present study, there were 6 cases (5.2%) with the CD56−CD16−
phenotype among 115 ENKTL cases. Prognosis analysis showed that the patients with the CD56−CD16− phenotype had a shorter OS than the patients with the CD56+CD16− phenotype (see Table 3). Maybe its earlier differentiation state correlates with the poor prognosis. We suggest that CD56− CD16− cases belong to the early CD56+CD16− mature stage, as the tumor cells usually contain granzyme B. Differentiation status is one prognostic factor in some malignancies of hematopoietic and lymphoid tissues [29]. For example, pre–T-like T-ALL with CD34 expression is associated with poorer survival compared with T-cell acute lymphoblastic leukemia (T-ALL) with later differentiation profiles [29]. However, van Grotel et al [30] reported that classification of pediatric T-ALL into T-cell developmental subgroups was not predictive of outcome. The CD56+CD16− subtype of ENKTL has phenotypic features similar to those of normal CD56brightCD16− NK cells [24-28]. We suggest that the tumor cells with the CD56+CD16− phenotype have higher proliferative ability than cells of the CD56dim/−CD16+ phenotype because all of the ENKTL cell lines originated from cells of the CD56+CD16− phenotype [18-20]. However, we identified a CD56dim/−CD16+ group with later differentiation status that was associated with a poor prognosis. Some cases of CD16+ ENKTL have been reported in several articles. For example, Suzuki et al [31] reported that
2262 11 (22.4%) of 49 cases expressed CD16 according to FCI, and 2 of them (18%) were located at an extranasal site. In the present study, we identified 19 CD16+ cases (16.5%), and 10 of them (52.6%) were located at extranasal sites. Some studies reported that aggressive NK-cell leukemia/lymphoma (ANKL) closely correlates with CD16 expression (more than 75%) and has a rapidly aggressive clinical course, but all of the ANKL cases were CD56+ [4,31,32]. We suggest that the CD56 dim/− CD16 + tumor cells of ENKTL mimic the CD56+CD16+ tumor cells of ANKL implying a higher aggressive ability. Furthermore, we observed that the CD56dim/−CD16+ group has a higher percentage (31.6%) of stage III/IV tumors than the other 2 groups (16.7% and 10%, respectively), and the CD56dim/−CD16+ group has a higher percentage (52.6%) of extranasal cases than the other 2 groups (0 and 15.6%, respectively), which may contribute to the poor prognosis. In conclusion, the current work demonstrates that the differentiation characteristics of these tumors are similar to the normal NK-cell developmental pattern, and 3 phenotypic subtypes were identified according to the expression of CD56 and CD16. The CD56dim/−CD16+ phenotypic subtype implies a poorer prognosis.
References [1] Jaffe ES, Harris NL, Stein H, Isaacson PG. Classification of lymphoid neoplasms: the microscope as a tool for disease discovery. Blood 2008; 112:4384-99. [2] Yang QP, Zhang WY, Yu JB, et al. Subtype distribution of lymphomas in Southwest China: analysis of 6,382 cases using WHO classification in a single institution. Diagn Pathol 2011;6:77. [3] Kim TM, Lee SY, Jeon YK, et al. Clinical heterogeneity of extranodal NK/T-cell lymphoma, nasal type: a national survey of the Korean Cancer Study Group. Ann Oncol 2008;19:1477-84. [4] Liang X, Graham DK. Natural killer cell neoplasms. Cancer 2008;112: 1425-36. [5] Chiang AK, Chan AC, Srivastava G, Ho FC. Nasal T/natural killer (NK)cell lymphomas are derived from Epstein-Barr virus-infected cytotoxic lymphocytes of both NK- and T-cell lineage. Int J Cancer 1997;73:332-8. [6] Jaffe ES. Nasal and nasal-type T/NK cell lymphoma: a unique form of lymphoma associated with the Epstein-Barr virus. Histopathology 1995;27:581-3. [7] Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 1997;3:730-7. [8] Cobaleda C, Sanchez-Garcia I. B-cell acute lymphoblastic leukaemia: towards understanding its cellular origin. Bioessays 2009;31:600-9. [9] Hope KJ, Jin L, Dick JE. Acute myeloid leukemia originates from a hierarchy of leukemic stem cell classes that differ in self-renewal capacity. Nat Immunol 2004;5:738-43. [10] Freud AG, Becknell B, Roychowdhury S, et al. A human CD34+ subset resides in lymph nodes and differentiates into CD56bright natural killer cells. Immunity 2005;22:295-304. [11] Huntington ND, Legrand N, Alves NL, et al. IL-15 trans-presentation promotes human NK cell development and differentiation in vivo. J Exp Med 2009;206:25-34. [12] Freud AG, Yokohama A, Becknell B, et al. Evidence for discrete stages of human natural killer cell differentiation in vivo. J Exp Med 2006;203:1033-43.
JB. Yu et al. [13] Romagnani C, Juelke K, Falco M, et al. CD56brightCD16− killer Ig-like receptor- NK cells display longer telomeres and acquire features of CD56dim NK cells upon activation. J Immunol 2007;178:4947-55. [14] Loza MJ, Perussia B. The IL-12 signature: NK cell terminal CD56+ high stage and effector functions. J Immunol 2004;172:88-96. [15] Cooper MA, Fehniger TA, Turner SC, et al. Human natural killer cells: a unique innate immunoregulatory role for the CD56 (bright) subset. Blood 2001;97:3146-51. [16] Ferlazzo G, Thomas D, Lin SL, et al. The abundant NK cells in human secondary lymphoid tissues require activation to express killer cell Ig-like receptors and become cytolytic. J Immunol 2004;172:1455-62. [17] Di Santo JP, Vosshenrich CA. Bone marrow versus thymic pathways of natural killer cell development. Immunol Rev 2006;214:35-46. [18] Nagata H, Konno A, Kimura N, et al. Characterization of novel natural killer (NK)-cell and gamma delta T-cell lines established from primary lesions of nasal T/NK-cell lymphomas associated with the EpsteinBarr virus. Blood 2001;97:708-13. [19] Tsuchiyama J, Yoshino T, Mori M, et al. Characterization of a novel human natural killer-cell line (NK-YS) established from natural killer cell lymphoma/leukemia associated with Epstein-Barr virus infection. Blood 1998;92:1374-83. [20] Kagami Y, Nakamura S, Suzuki R, et al. Establishment of an IL-2dependent cell line derived from “nasal-type” NK/T-cell lymphoma of CD2+, sCD3−, CD3epsilon+, CD56+ phenotype and associated with the Epstein-Barr virus. Br J Haematol 1998;103:669-77. [21] Zhao S, Tang QL, He MX, et al. A novel nude mice model of human extranodal nasal type NK/T-cell lymphoma. Leukemia 2008;22: 170-8. [22] Yu JB, Zuo Z, Tang Y, et al. Extranodal nasal-type natural killer/T-cell lymphoma of the skin: a clinicopathologic study of 16 cases in China. HUM PATHOL 2009;40:807-16. [23] Craig FE, Foon KA. Flow cytometric immunophenotyping for hematologic neoplasms. Blood 2008;111:3941-67. [24] Bugalia A, Manipadam MT, Nair S. Immunomorphologic profile and Epstein-Barr virus status of a cohort of 35 cases of extranodal natural killer/T-cell lymphoma, nasal type of upper aerodigestive tract from a tertiary care center in South India. Leuk Lymphoma 2013;54:1201-7. [25] Lu D, Lin CN, Chuang SS, Hwang WS, Huang WT. T-cell and NK/T-cell lymphomas in southern Taiwan: a study of 72 cases in a single institute. Leuk Lymphoma 2004;45:923-8. [26] Gaal K, Sun NC, Hernandez AM, Arber DA. Sinonasal NK/T-cell lymphomas in the United States. Am J Surg Pathol 2000;24:1511-7. [27] Ko YH, Ree HJ, Kim WS, Choi WH, Moon WS, Kim SW. Clinicopathologic and genotypic study of extranodal nasal-type natural killer/T-cell lymphoma and natural killer precursor lymphoma among Koreans. Cancer 2000;89:2106-16. [28] Kim JE, Kim YA, Jeon YK, Park SS, Heo DS, Kim CW. Comparative analysis of NK/T-cell lymphoma and peripheral Tcell lymphoma in Korea: clinicopathological correlations and analysis of EBV strain type and 30-bp deletion variant LMP1. Pathol Int 2003;53:735-43. [29] Babusikova O, Stevulova L, Fajtova M. Immunophenotyping parameters as prognostic factors in T-acute leukemia patients. Neoplasma 2009;56:508-13. [30] van Grotel M, Meijerink JP, van Wering ER, et al. Prognostic significance of molecular-cytogenetic abnormalities in pediatric T-ALL is not explained by immunophenotypic differences. Leukemia 2008;22:124-31. [31] Suzuki R, Suzumiya J, Yamaguchi M, et al. Prognostic factors for mature natural killer (NK) cell neoplasms: aggressive NK cell leukemia and extranodal NK cell lymphoma, nasal type. Ann Oncol 2010;21:1032-40. [32] Ishida F, Ko YH, Kim WS, et al. Aggressive natural killer cell leukemia: therapeutic potential of L-asparaginase and allogeneic hematopoietic stem cell transplantation. Cancer Sci 2012;103:1079-83.
www.elsevier.com/locate/humpath
Original contribution
Identification of immunophenotypic subtypes with different prognoses in extranodal natural killer/T-cell lymphoma, nasal type☆,☆☆ Jian-Bo Yu PhD a,b,1 , Zhuo Zuo PhD a,1 , Wen-Yan Zhang PhD a , Qun-Pei Yang PhD a , Ying-Chun Zhang PhD a , Yuan Tang PhD a , Sha Zhao PhD a , Xian-Ming Mo PhD a , Wei-Ping Liu MD a,⁎ a
Department of Pathology and Laboratory of Stem Cell Biology, State Key Laboratory of Biotherapy, West China Hospital of Sichuan University, Chengdu, 610041, China b Department of Pathology, Key Laboratory of Cancer Prevention and Treatment of Heilongjiang Province, Mudanjiang Medical University, Mudanjiang, 157011, China Received 14 March 2014; revised 21 April 2014; accepted 24 April 2014
Keywords: CD16; CD56; Lymphoma; Natural killer cells; T lymphocytes
Summary To analyze the differentiation characteristics of extranodal natural killer/T-cell lymphoma, nasal type, one nude mouse model, cell lines SNK6 and SNT8, and 16 fresh human samples were analyzed by flow cytometry immunophenotyping and immunohistochemistry staining; and 115 archived cases were used for phenotypic detection and prognostic analysis. We found that CD25 was expressed by most tumor cells in all samples, and CD56+CD25+ cells were the predominant population in the mouse model, the 2 cell lines, and 10 of the 16 fresh tumor samples; in the other 6 fresh tumor samples, the predominant cell population was of the CD16+CD25+ phenotype, and only a minor population showed the CD56+CD25+ phenotype. The phenotype detected by immunohistochemistry staining generally was consistent with the phenotype found by flow cytometry immunophenotyping. According to the expression of CD56 and CD16, 115 cases could be classified into 3 phenotypic subtypes: CD56−CD16−, CD56+CD16−, and CD56dim/−CD16+. Patients with tumors of the CD56dim/−CD16+ phenotype had a poorer prognosis than patients with tumors of the other phenotypes. Differentiation of extranodal natural killer/T-cell lymphoma, nasal type apparently resembles the normal natural killer cell developmental pattern, and these tumors can be classified into 3 phenotypic subtypes of different aggressiveness. Expression of CD56dim/−CD16+ implies a poorer prognosis. © 2014 Elsevier Inc. All rights reserved.
☆ Competing interests: The authors have declared no potential conflicts of interest. ☆☆ Funding/Support: This work was supported by The National Natural Science Foundation of China (Beijing) (no. 81272626 and 81271628) and the Innovation Team Project of Tumor Prevention and Treatment in Heilongjiang University, Health Department Foundation of Heilongjiang Province (no. 2009-412). ⁎ Corresponding author. Department of Pathology, West China Hospital of Sichuan University, No. 37, Guoxue Alley, Chengdu 610041 Sichuan, China. E-mail address: [email protected] (W. -P. Liu). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.humpath.2014.04.024 0046-8177/© 2014 Elsevier Inc. All rights reserved.
1. Introduction Human extranodal natural killer (NK)/T-cell lymphoma, nasal type (ENKTL), is a distinct clinicopathological entity associated with a highly aggressive clinical course and thus
2256
JB. Yu et al. ative support for investigations of the differentiation patterns of ENKTL. In human bone marrow and secondary lymphoid tissue (SLT), CD34+ hematopoietic stem cells and hematopoietic progenitor cells can give rise to cytolytic NK cells after stimulation with either interleukin (IL)-2 or IL-15 [10,11]. Freud et al [12] used the combination of CD34, CD117, and CD94 antigens to discriminate 4 functionally distinct stages of human CD56bright NK-cell development within SLT. Some reports demonstrated that both CD56 bright CD16 − and CD56dim/−CD16bright cells are mature NK cells, whereas CD56brightCD16− NK cells could be the direct precursors of CD56dim/−CD16bright NK cells [13-17]. Some authors believe that CD56brightCD16− and CD56dim/−CD16bright NK cells are derived from SLT and bone marrow, respectively [14-17]. Whether the CD56brightCD16− and CD56dim/−CD16bright phenotypes occur in ENKTL remains uncertain. In this study, based on reports describing normal NKcell development, differentiation patterns of ENKTL were investigated. The samples included a nude mouse model, 2 ENKTL cell lines, 16 fresh tissue samples, and 115 paraffin-embedded block samples from human ENKTL. Immunophenotypic analysis by both flow cytometry immunophenotyping (FCI) and immunohistochemistry (IHC) staining demonstrated that tumor-cell populations in ENKTL represent a continuum of normal stages of
a poor prognosis [1]. It is rare in Western countries but more common in Asia, Mexico, and Central and South America [1,2]. The clinical features are more aggressive, and the prognosis is poorer in extranasal than in nasal tumors [3]. Most of the tumors exhibit NK-cell phenotypes with expression of CD56 and cytoplasmic CD3 but do not express surface CD3; and T-cell receptor (TCR) genes remain in the germline configuration [4]. However, some cases have a clonal T-cell phenotype and genotype [5,6]. The differentiation patterns of lymphatic and hematopoietic leukemia have been well described, but the differentiation patterns of lymphomas, including ENKTL, are less clear. A series of articles reported that hematopoietic stem cells and committed progenitor cells could transform into tumorinitiating cells and differentiate along the original lineage [7]. B-cell acute lymphoid leukemia is phenotypically reminiscent of the normal stages of B-cell differentiation, and a scheme that assigns stages of human B-cell acute lymphoid leukemia similar to those of normal B cells has been proposed [8]. This developmental pattern suggests that leukemia represents normal hematopoietic development gone slightly awry [9]. Some studies have elucidated the process of normal NK-cell development [10-17]. These studies provide comparTable 1
Clinicopathological features and immunophenotype of 16 human tumor samples, nude mouse model, and cell lines of ENKTL
Sample no.
Human tumor 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Mouse model SNK6 [18] SNT8 [18] NK-YS [19] HANK-1 [20]
Sex/ age
Primary lesion
F/42 M/42 M/39 F/35 M/59 M/18 F/53 M/33 M/33 M/19 M/63 F/51 F/25 F/62 M/37 M/42 M/48 M/62 F/48 F/19 F/46
Nasal Nasal Nasal Skin Colon Palate Palate Nasal Palate Nasal Nasal Colon Palate Nasal Pharynx Skin Nasal Nasal Nasal Nasal Retroperitoneal
IHC
FCI
CD3ε
CD20
CD56
GB
+ + + + + + + + + + + + + + + + + + + + +
− − − − − − − − − − − − − − − − − − − − NA
+ + − + + + + + − − − + − − + + + + + + +
+ + + + + + + + + + + + + + + + + NA NA NA +
CD56+CD16− (%) 67 87 22 67 39 88 66 93 6 5 15 90 12 17 64 89 90 86 76 + +
CD56dim/−CD16+ (%) NA NA 54 0 0 NA NA 0 90 91 70 0 82 33 0 0 0 0 0 0 0
TCR rearr. − − − − + NA NA − + − − − − − NA − − − + − −
NOTE. All specimens were positive for EBV. The percentage of CD56+CD16− and CD56dim/-CD16+ cells was calculated under CD25+ gating; SNK6, SNT8, NK-YS, and HANK-1 are ENKTL cell lines. Only flow cytometry analysis was done in the present study; the other information is from the literature. All of the information on NK-YS and HANK-1 is from the literature. Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; IHC, immunohistochemistry; FCI, flow cytometry immunophenotyping; NA, not available; GB, granzyme B; rearr., rearrangement.
Phenotypic subtypes in ENKTL human NK-cell differentiation; and CD56 − CD16 − , CD56+ CD16− , and CD56 dim/− CD16+ phenotypic subtypes were identified. The CD56 dim/− CD16 + subtype was associated with a poorer prognosis.
2. Materials and methods 2.1. Patients Sixteen fresh samples of human ENKTL were collected from West China Hospital of Sichuan University for FCI analysis (Table 1). Before collection of these samples, none of the patients had received any antitumor treatment. Informed consent was obtained from all enrolled patients, and the procedure for tissue collection was approved by the Human Research Ethical Committees of West China Hospital, Sichuan University. In addition, archived samples harvested from 115 patients with ENKTL between January 1993 and December 2010 were obtained from the Pathology Department of West China Hospital and used for clinicopathological review, IHC staining, and prognostic analysis. The diagnosis of all cases was made by 2 experienced hematopathologists using the updated World Health Organization classification of tumors of the hematopoietic and lymphoid tissues [1].
2.2. Nude mouse tumor model and tumor cell lines A BALB/c nude mouse model of human ENKTL was established using secondary gastric NK/T-cell lymphoma and passaged as described previously [21]. Two ENKTL cell lines, SNK6 and SNT8, were kindly provided by Prof Nagata [18]. The SNK6 cells expressed the phenotype CD3−CD4− CD8−CD56+, and TCRγ genes were not rearranged; SNT8 expressed the phenotype CD3 + CD4 − CD8 − CD56 + and exhibited TCRγ rearrangement. Both SNK6 and SNT8 were cultured in RPMI 1640 (Gibco, Grand Island, NY) medium supplemented with antibiotics, 10% heat-inactivated human serum, and recombinant human IL-2 600 U/mL and were incubated in a humidified atmosphere of 5% carbon dioxide at 37°C.
2.3. Cell preparation Neoplastic tissues from human ENKTL or from the BALB/c nude mice were cut into thin slices, suspended in RPMI 1640 medium, and digested using collagenase 3 (Worthington Biochemical, Lakewood, NJ). The mixture was incubated at 37°C for as long as 3 hours to allow complete digestion. Cells were filtered through 40-μm nylon mesh and washed twice with phosphate-buffered saline (PBS) supplemented with 2% fetal calf serum and 0.1% sodium azide and then stained for FCI analysis.
2257
2.4. Antibody staining for FCI Suspensions of mononuclear cells prepared by tissue digestion were incubated with antibodies for 20 minutes at 4°C, washed with PBS, resuspended in PBS, and analyzed by FCI. Fluorochrome-conjugated monoclonal antibodies against the following antigens were purchased from PharMingen (Becton Dickinson, San Diego, CA): CD3 (UCHT1), CD4 (RPA-T4), CD7 (M-T701), CD8 (SK1), CD14 (M5E2), CD16 (3G8), CD20 (2H7), CD25 (M-A251), CD31 (WM59), CD34 (581), CD38 (HIT2), CD44 (G4426), CD45 (H130), CD56 (B159), CD94 (HP-3D9), CD117 (YB5.B8), CD122 (Mik-β2), HLA-DR (TU36), integrin β7 (FIB504), and isotype control immunoglobulin G (R3-34). Dead cells were distinguished by staining with 7-amino-actinomycin D. All fresh samples and cell lines were gated according to their light scatter characteristics. Data were analyzed using FACS/Calibour CELLQuest, Becton Dickinson (San Jose, CA).
2.5. IHC staining Antibodies used for IHC included CD3ε (polyclonal; Dako, Glostrup, Denmark), CD16 (275003; R&D, Minneapolis, MN), CD20 (L26; Dako), CD25 (IL2R.1; Zymed, Grand Island, NY), CD34 (QBEnd/10; NeoMarker, Fremont, CA), CD45RO (UCHL1; Dako), CD56 (123, C3; Zymed), granzyme B (GZB01; NeoMarkers), and CD117 (2E4; Zymed). Sections (4 μm) were cut and deparaffinized in xylene and then hydrated in a graded series of ethanol. The slides were treated by pressure cooking in citric acid buffer (10 mmol/L, pH 7.4) for 3 minutes before staining. Bound antibodies were visualized by the EnVision (Dako) method using 3,3′-diaminobenzidine as a substrate.
2.6. In situ hybridization and polymerase chain reaction The in situ hybridization (ISH) was carried out with a fluorescein-labeled oligonucleotide probe complementary to 2 Epstein-Barr virus (EBV)–Epstein-Barr virus encoded small RNAs 1 and 2 (EBER1/2) (Y520001; Dako). Rabbit antifluorescein isothiocyanate (FITC) antibody conjugated with alkaline phosphatase (Dako) was used to detect the probe. Genomic DNA from paraffin-embedded tissue samples was extracted by phenol-chloroform procedures. Both universal TCRγ (TVG/TJX) primers and BIOMED-2 polymerase chain reaction (PCR) multiplex tubes (TCRβ + TCRγ) were used for TCR rearrangement analysis by PCR. These procedures were performed as we previously reported [22].
2.7. Survival and statistical analysis Follow-up data were available for all 115 patients. The overall survival (OS) time was plotted using the Kaplan-Meier
2258 procedure (SPSS 13.0; SPSS, Chicago, IL). Differences in survival times between patient groups were calculated using the log-rank test. P b .05 was considered statistically significant.
3. Results 3.1. Differentiation characteristic detected by FCI in ENKTL For discovering the differentiation characteristics of ENKTL, a nude mouse tumor model, 2 cell lines, and 16 samples of fresh human ENKTL were used for immunophenotypic analysis by FCI using a series of antibodies. We found that CD25 was expressed by a majority of the tumor cells in all samples (Fig. 1A), and some of these cells coexpressed CD56, CD117, and CD122. We also observed some CD117+ cells also expressing CD56. In the mouse model and 2 cell lines, CD25+ CD56+ cells were the predominant population (Fig. 1B), and CD25+CD56− cells were a minor population. In the SNK6 and SNT8 tumor cell lines, the phenotype features of the lines changed during culture: after 24 hours, CD56+ tumor cells represented 30% of the total, but this figure increased to more than 76% by 72 hours (Fig. 1C and D), and CD34+ cells were observed consistently in small numbers (Fig. 1E). However, when we analyzed 16 fresh human tumor samples, a predominance of CD56+CD25+ cells was observed in only 10, and there was no CD16 expression in these cases (Fig. 1F). In the other 6 samples (see Table 1), a lower frequency of CD56
JB. Yu et al. expression was noticed, and further analysis revealed that the phenotype of the predominant cells was CD16+ (Fig. 1G). In the CD16+ samples, a minor CD56+CD16− cell population could be observed, but no CD20+ or CD14+ cells were present. The CD117 (c-kit) protein typically is expressed by precursor and progenitor cells, and CD122 is the IL-2/IL-15 receptor β chain and a marker for immature NK cells [10]. Both were detected in minor cell populations in all samples (Fig. 1E and H). Expression of CD94 was low, and a small percentage of CD3+ tumor cells was observed in cases 5 and 9. CD8+ tumor cells were noted in cases 5, 9, 10, 11, and 14. No CD4 expression was observed in any cells.
3.2. Immunophenotype characteristics detected by IHC in ENKTL We also assessed phenotypic characteristics using CD56 and CD16 markers detected by IHC (Fig. 2). Expression of CD56 and CD16 generally coincided with the phenotypes found by FCI. However, when the case with no more than 22% of CD56dim tumor cells detected by FCI, only weak or no CD56 expression was observed by IHC (see Table 1). This might be the reason that a low density of CD56 antigen on tumor cells is difficult to detect using the IHC method. Based on the expression characteristics of the CD56 and CD16 markers, the tumors could be classified into 3 phenotypic subtypes. The first, no matter weak or no CD56 expression (CD56dim/− ), the CD16+ cases were classified as the CD56dim/−CD16+ phenotype, similar to normal NK cells [13-15]. The second was CD56+ cases without CD16 expression, which was classified as the
Fig. 1 Immunophenotypic features of ENKTL detected by FCI. Dead cells were excluded by 7-amino-actinomycin D staining. Dot plots were gated based on light scatter characteristics and isotype controls. A, Most tumor cells in the nude mouse model were CD25+ and CD16−. B, CD56+CD25+ tumor cells were the predominant population in the nude mouse model. C, After 24 hours in culture, a minority of proliferating CD56+ cells were observed in SNK6. D, A majority of cells expressing CD56 were observed after 72 hours in culture. E, A minimum of CD34+ and CD122+ cells were observed in the 2 cell lines. F, No CD16+ cells were observed in 10 of 16 fresh tumor samples. G, Of 16 fresh tumor samples, 6 were CD16+, and a fraction of the cells expressed the CD56+CD16− phenotype. H, CD117+ tumor cells were observed in all samples, some of these cells were CD25−.
Phenotypic subtypes in ENKTL
2259
Fig. 2 Immunophenotypic features detected by IHC. A to C, Tumor cells with CD56+CD16− phenotype. A, Strong expression of CD56. B, Cells are negative for CD16. C, Granzyme B was moderately expressed. D to F, CD56dim/−CD16+ cases. D, No expression of CD56. E, Expression of CD16. F, Strong expression of granzyme B. Original magnification, × 200.
CD56+CD16− phenotype. The third was CD56− cases without CD16 expression, which was classified as a CD56−CD16− phenotype. The CD25, CD34, and CD117 markers also were sought by IHC, but no positive tumor cells were observed. Expression of granzyme B was observed in all 3 phenotypes with no significant difference among the phenotypes. In addition, EBER1/2 staining by ISH was positive in all cases. Case 5 (CD56+CD16− phenotype) and case 9 (CD56dim/− CD16+ phenotype) contained TCRγ gene rearrangements. No other cells expressed this feature.
3.3. Clinicopathological characteristics and prognostic analysis of 3 subtypes of ENKTL Archived data of 115 ENKTL cases were reviewed (Table 2). The patients included 87 male and 28 female subjects with a median age of 40 years (range, 6-84 years). Forty-one patients (35.7%) had B symptoms. Ninety-one patients (79.1%) had tumors in the nasal/upper aerodigestive tract, and 24 (20.9%) presented with extranasal sites of involvement as follows: small intestine (n = 5), colon (n = 8), skin (n = 8), and lymph nodes (n = 3). Sixteen patients (13.9%) were in stage III/IV. Six patients (5.2%) had tumors at more than one site. All cases showed coagulative necrosis to various degrees. Angiocentric infiltration was observed in 15 cases (13%). The tumors showed predominantly medium-sized cells (n = 105 [91.3%]). In 10 cases (8.7%), the tumor cells were predominantly large. Six patients were treated with radiation therapy, 24 patients with chemotherapy, 15 patients with operation and chemotherapy, and 4 patients with antiinflammatory therapy; and 45 patients received combined radiation
and chemotherapy. The most commonly used chemotherapy regimen was cyclophosphamide, doxorubicin, vincristine, and prednisolone. Treatment information was not available for 21 patients. Follow-up data were available for all 115 patients (range, 0.08-5.0 years). With a median follow-up of 0.92 years, 59 patients (51.3%) had died. The median OS time was 1.31 years, and the OS was 48.7%. The results of IHC are summarized in Table 2. All cases were CD20−. CD3ε was positive in 100 of 102 cases, granzyme B was positive in 110 of 111 cases, CD56 was positive in 90 of 115 cases, and CD16 was positive in 19 of 115 cases. EBV-positive cells were detected by ISH in all 115 cases. According to the expression of CD56 and CD16, 115 cases were classified into 3 phenotypic groups, 6 CD56− CD16−, 90 CD56+CD16−, and 19 CD56dim/−CD16+. Follow-up data were used for prognosis analysis based on phenotypic subtypes, B symptom, lesions, lesion sites, and stages (Table 3, Fig. 3). The OS for the CD56+CD16− group was 57.8% compared with 33.3% for the CD56−CD16− group and 10.5% for the CD56dim/−CD16+ group (P b .0001). The patients with B symptoms had a poorer prognosis, although the difference was not statistically significant (P = .198). Stage III/IV disease, more than one lesion, and extranasal lesions correlated significantly with a poorer prognosis (P = .002, P = .0003, and P b .0001, respectively).
4. Discussion The ENKTL is the most common non–B-cell extranodal neoplasm seen in China, representing approximately 17% of all non-Hodgkin lymphomas [2]. This study was based on
2260 Table 2
JB. Yu et al. Clinicopathological data of 115 cases of ENKTL CD56−CD16−, n (%)
CD56+CD16−, n (%)
CD56dim/−CD16+, n (%)
115
6 (5.2)
90 (78.3)
19 (16.5)
87 (75.7) 28 (24.3)
5 (83.3) 1 (16.7)
67 (74.4) 23 (25.6)
15 (78.9) 4 (21.1)
40 6-84
38 25-60
43 15-84
36 6-69
41 (35.7) 74 (64.3)
1 (16.7) 5 (83.3)
31 (34.4) 59 (65.6)
9 (47.4) 10 (52.6)
10 (8.7) 105 (91.3)
1 (16.7) 5 (83.3)
8 (8.9) 82 (91.1)
1 (5.3) 18 (94.7)
15 (13) 100 (87)
1 (16.7) 5 (83.3)
10 (11.1) 80 (88.9)
4 (21.1) 15 (78.9)
100/102 (98.0) 110/111 (99.1) 0 115 (100)
4/4 (100) 6 (100) 0 6 (100)
79/81 (97.5) 86/86 (100) 0 90 (100)
17/17 (100) 18/19 (94.7) 0 19 (100)
99 (86.1) 16 (13.9)
5 (83.3) 1 (16.7)
81 (90) 9 (10)
13 (68.4) 6 (31.6)
91 (79.1) 24 (20.9)
6 (100) 0
76 (84.4) 14 (15.6)
9 (47.4) 10 (52.6)
109 (94.8) 6 (5.2%)
5 (83.3) 1 (16.7)
88 (97.8) 2 (2.2)
16 (84.2) 3 (15.8)
59 (51.3) 56 (48.7)
4 (66.7) 2 (33.3)
38 (42.2) 52 (57.8)
17 (89.5) 2 (10.5)
Total, n (%) Total no. Sex (%) Male Female Age (y) Median Range B symptoms (%) Yes No Cell size (%) Large Medium Angiocentric infiltration (%) Yes No IHC and EBV (%) CD3ε Granzyme B CD20 EBV Stage (%) I/II III/IV Lesion site (%) Nasal Extranasal Lesion size (cm) (%) 1 ≥2 Last known condition (%) Dead Alive
Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; IHC, immunohistochemistry; EBV, Epstein-Barr virus.
recent advances in delineating normal NK-cell development [10-17]. The expression of CD117 and CD34 marks precursor cells; CD56 expression in NK-cell development shows a dynamic process, which increases gradually from the immature stage to the CD56brightCD16− mature stage and then gradually wanes to CD56dim/−CD16+ [11,12]. In the present study, we observed minor cell populations with the CD117+ and CD34+ phenotypes in ENKTL, and CD56 expression gradually increased, as the duration of culture of SNK6 and SNT8 cells was extended. A minor percentage of the CD56+ tumor cell population was found in the CD56dim/− CD16+ phenotypic cases. These facts support the conclusion that the development process of the tumor represents a continuum of normal NK cells, even in the cases accompanied by TCRγ gene rearrangement; and the terminal point of tumor cell differentiation in different phenotypic subtypes may cease at different points of the developmental process from immature to mature cells. Craig and Foon [23] regarded flow cytometry as a valuable tool for identifying phenotypes and lineages and for the
Table 3
Prognostic analysis of 115 cases of ENKTL
B symptoms Yes No Lesions Nasal Extranasal Lesion size (cm) 1 ≥2 Stage I/II III/IV Phenotypic subtype CD56−CD16− CD56+CD16− CD56dim/−CD16+
n (%)
OS, n (%)
41 74
17 (41.5) 39 (52.7)
91 24
54 (59.3) 2 (8.3)
109 6
56 (51.4) 0
99 16
53 (53.5) 3 (18.8)
6 90 19
2 (33.3) 52 (57.8) 2 (10.5)
P .198
b.0001
.0003
.002
b.0001
Abbreviations: ENKTL, extranodal natural killer/T-cell lymphoma, nasal type; OS, overall survival.
Phenotypic subtypes in ENKTL
2261
Fig. 3 Prognosis analysis. OS analysis was performed on 115 cases with complete follow-up data. A, B symptoms were associated with shorter OS but without statistical significance. B, Patients whose tumors had the CD56dim/−CD16+ phenotype had poorer OS, whereas the CD56+CD16− phenotype was associated with a better OS. The results of prognosis analysis showed that stage III/IV (C), more than one lesion (D), and extranasal tumors (E) were significantly associated with a poorer prognosis.
diagnosis and evaluation of hematologic neoplasms. The CD25 protein is the IL-2 receptor α chain. Its ligand, IL-2, is an important cytokine for normal NK-cell growth [10-12]. In the present study, CD25 expression detected by FCI demonstrated that IL-2 is a key cytokine for ENKTL and the cell lines arising from it [18-20]. Furthermore, the immunophenotype detected by FCI demonstrated that not all CD56− cases were authentic T-cell–lineage tumors [24]. When CD56dim tumor cells were detected by FCI, the CD56 marker was difficult to detect by IHC, and only 1 of 6 CD56− cases (case 9) exhibited TCRγ rearrangement. Our finding indicates that CD56− cases in ENKTL might mimic normal NK cells with CD56dim/−CD16+ and CD56−/dimCD16− differentiation. Therefore, according to the expression of CD56 and CD16 detected by IHC, we suggest that when CD56− cases are accompanied by CD16 expression, the differentiation pathway of the tumor cells may pass through a CD56+CD16− stage on their way to the terminal CD56dim/− CD16+ stage. When the CD56 marker was expressed, the differentiation of the tumor cells may cease at the CD56+CD16− mature stage; when CD56− cases did not manifest CD16 expression, differentiation may have ceased at the early CD56+CD16− mature stage or a later immature stage. Tumors not expressing CD56 or CD16 have been recognized for more than a decade [4,5,25-28]. In the present study, there were 6 cases (5.2%) with the CD56−CD16−
phenotype among 115 ENKTL cases. Prognosis analysis showed that the patients with the CD56−CD16− phenotype had a shorter OS than the patients with the CD56+CD16− phenotype (see Table 3). Maybe its earlier differentiation state correlates with the poor prognosis. We suggest that CD56− CD16− cases belong to the early CD56+CD16− mature stage, as the tumor cells usually contain granzyme B. Differentiation status is one prognostic factor in some malignancies of hematopoietic and lymphoid tissues [29]. For example, pre–T-like T-ALL with CD34 expression is associated with poorer survival compared with T-cell acute lymphoblastic leukemia (T-ALL) with later differentiation profiles [29]. However, van Grotel et al [30] reported that classification of pediatric T-ALL into T-cell developmental subgroups was not predictive of outcome. The CD56+CD16− subtype of ENKTL has phenotypic features similar to those of normal CD56brightCD16− NK cells [24-28]. We suggest that the tumor cells with the CD56+CD16− phenotype have higher proliferative ability than cells of the CD56dim/−CD16+ phenotype because all of the ENKTL cell lines originated from cells of the CD56+CD16− phenotype [18-20]. However, we identified a CD56dim/−CD16+ group with later differentiation status that was associated with a poor prognosis. Some cases of CD16+ ENKTL have been reported in several articles. For example, Suzuki et al [31] reported that
2262 11 (22.4%) of 49 cases expressed CD16 according to FCI, and 2 of them (18%) were located at an extranasal site. In the present study, we identified 19 CD16+ cases (16.5%), and 10 of them (52.6%) were located at extranasal sites. Some studies reported that aggressive NK-cell leukemia/lymphoma (ANKL) closely correlates with CD16 expression (more than 75%) and has a rapidly aggressive clinical course, but all of the ANKL cases were CD56+ [4,31,32]. We suggest that the CD56 dim/− CD16 + tumor cells of ENKTL mimic the CD56+CD16+ tumor cells of ANKL implying a higher aggressive ability. Furthermore, we observed that the CD56dim/−CD16+ group has a higher percentage (31.6%) of stage III/IV tumors than the other 2 groups (16.7% and 10%, respectively), and the CD56dim/−CD16+ group has a higher percentage (52.6%) of extranasal cases than the other 2 groups (0 and 15.6%, respectively), which may contribute to the poor prognosis. In conclusion, the current work demonstrates that the differentiation characteristics of these tumors are similar to the normal NK-cell developmental pattern, and 3 phenotypic subtypes were identified according to the expression of CD56 and CD16. The CD56dim/−CD16+ phenotypic subtype implies a poorer prognosis.
References [1] Jaffe ES, Harris NL, Stein H, Isaacson PG. Classification of lymphoid neoplasms: the microscope as a tool for disease discovery. Blood 2008; 112:4384-99. [2] Yang QP, Zhang WY, Yu JB, et al. Subtype distribution of lymphomas in Southwest China: analysis of 6,382 cases using WHO classification in a single institution. Diagn Pathol 2011;6:77. [3] Kim TM, Lee SY, Jeon YK, et al. Clinical heterogeneity of extranodal NK/T-cell lymphoma, nasal type: a national survey of the Korean Cancer Study Group. Ann Oncol 2008;19:1477-84. [4] Liang X, Graham DK. Natural killer cell neoplasms. Cancer 2008;112: 1425-36. [5] Chiang AK, Chan AC, Srivastava G, Ho FC. Nasal T/natural killer (NK)cell lymphomas are derived from Epstein-Barr virus-infected cytotoxic lymphocytes of both NK- and T-cell lineage. Int J Cancer 1997;73:332-8. [6] Jaffe ES. Nasal and nasal-type T/NK cell lymphoma: a unique form of lymphoma associated with the Epstein-Barr virus. Histopathology 1995;27:581-3. [7] Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med 1997;3:730-7. [8] Cobaleda C, Sanchez-Garcia I. B-cell acute lymphoblastic leukaemia: towards understanding its cellular origin. Bioessays 2009;31:600-9. [9] Hope KJ, Jin L, Dick JE. Acute myeloid leukemia originates from a hierarchy of leukemic stem cell classes that differ in self-renewal capacity. Nat Immunol 2004;5:738-43. [10] Freud AG, Becknell B, Roychowdhury S, et al. A human CD34+ subset resides in lymph nodes and differentiates into CD56bright natural killer cells. Immunity 2005;22:295-304. [11] Huntington ND, Legrand N, Alves NL, et al. IL-15 trans-presentation promotes human NK cell development and differentiation in vivo. J Exp Med 2009;206:25-34. [12] Freud AG, Yokohama A, Becknell B, et al. Evidence for discrete stages of human natural killer cell differentiation in vivo. J Exp Med 2006;203:1033-43.
JB. Yu et al. [13] Romagnani C, Juelke K, Falco M, et al. CD56brightCD16− killer Ig-like receptor- NK cells display longer telomeres and acquire features of CD56dim NK cells upon activation. J Immunol 2007;178:4947-55. [14] Loza MJ, Perussia B. The IL-12 signature: NK cell terminal CD56+ high stage and effector functions. J Immunol 2004;172:88-96. [15] Cooper MA, Fehniger TA, Turner SC, et al. Human natural killer cells: a unique innate immunoregulatory role for the CD56 (bright) subset. Blood 2001;97:3146-51. [16] Ferlazzo G, Thomas D, Lin SL, et al. The abundant NK cells in human secondary lymphoid tissues require activation to express killer cell Ig-like receptors and become cytolytic. J Immunol 2004;172:1455-62. [17] Di Santo JP, Vosshenrich CA. Bone marrow versus thymic pathways of natural killer cell development. Immunol Rev 2006;214:35-46. [18] Nagata H, Konno A, Kimura N, et al. Characterization of novel natural killer (NK)-cell and gamma delta T-cell lines established from primary lesions of nasal T/NK-cell lymphomas associated with the EpsteinBarr virus. Blood 2001;97:708-13. [19] Tsuchiyama J, Yoshino T, Mori M, et al. Characterization of a novel human natural killer-cell line (NK-YS) established from natural killer cell lymphoma/leukemia associated with Epstein-Barr virus infection. Blood 1998;92:1374-83. [20] Kagami Y, Nakamura S, Suzuki R, et al. Establishment of an IL-2dependent cell line derived from “nasal-type” NK/T-cell lymphoma of CD2+, sCD3−, CD3epsilon+, CD56+ phenotype and associated with the Epstein-Barr virus. Br J Haematol 1998;103:669-77. [21] Zhao S, Tang QL, He MX, et al. A novel nude mice model of human extranodal nasal type NK/T-cell lymphoma. Leukemia 2008;22: 170-8. [22] Yu JB, Zuo Z, Tang Y, et al. Extranodal nasal-type natural killer/T-cell lymphoma of the skin: a clinicopathologic study of 16 cases in China. HUM PATHOL 2009;40:807-16. [23] Craig FE, Foon KA. Flow cytometric immunophenotyping for hematologic neoplasms. Blood 2008;111:3941-67. [24] Bugalia A, Manipadam MT, Nair S. Immunomorphologic profile and Epstein-Barr virus status of a cohort of 35 cases of extranodal natural killer/T-cell lymphoma, nasal type of upper aerodigestive tract from a tertiary care center in South India. Leuk Lymphoma 2013;54:1201-7. [25] Lu D, Lin CN, Chuang SS, Hwang WS, Huang WT. T-cell and NK/T-cell lymphomas in southern Taiwan: a study of 72 cases in a single institute. Leuk Lymphoma 2004;45:923-8. [26] Gaal K, Sun NC, Hernandez AM, Arber DA. Sinonasal NK/T-cell lymphomas in the United States. Am J Surg Pathol 2000;24:1511-7. [27] Ko YH, Ree HJ, Kim WS, Choi WH, Moon WS, Kim SW. Clinicopathologic and genotypic study of extranodal nasal-type natural killer/T-cell lymphoma and natural killer precursor lymphoma among Koreans. Cancer 2000;89:2106-16. [28] Kim JE, Kim YA, Jeon YK, Park SS, Heo DS, Kim CW. Comparative analysis of NK/T-cell lymphoma and peripheral Tcell lymphoma in Korea: clinicopathological correlations and analysis of EBV strain type and 30-bp deletion variant LMP1. Pathol Int 2003;53:735-43. [29] Babusikova O, Stevulova L, Fajtova M. Immunophenotyping parameters as prognostic factors in T-acute leukemia patients. Neoplasma 2009;56:508-13. [30] van Grotel M, Meijerink JP, van Wering ER, et al. Prognostic significance of molecular-cytogenetic abnormalities in pediatric T-ALL is not explained by immunophenotypic differences. Leukemia 2008;22:124-31. [31] Suzuki R, Suzumiya J, Yamaguchi M, et al. Prognostic factors for mature natural killer (NK) cell neoplasms: aggressive NK cell leukemia and extranodal NK cell lymphoma, nasal type. Ann Oncol 2010;21:1032-40. [32] Ishida F, Ko YH, Kim WS, et al. Aggressive natural killer cell leukemia: therapeutic potential of L-asparaginase and allogeneic hematopoietic stem cell transplantation. Cancer Sci 2012;103:1079-83.