Vol. 117 No. 3 March 2014
The clinicopathologic significance of the expression of HLA-G in oral squamous cell carcinoma Andréia Souza Gonçalves, DDS,a Isabela Jubé Wastowski, BS, PhD,b Lucas Raineri Capeletti,c Nancy Tomoko Sacono, DDS, PhD,d Alane Pereira Cortez, BS, PhD,e Marize Campos Valadares, BS, PhD,f Tarcília Aparecida Silva, DDS, PhD,g and Aline Carvalho Batista, DDS, PhDh Federal University of Goiás, Goiânia, Brazil; State University of Goiás, Morrinhos, Brazil; Pontifical Catholic University of Goiás, Goiânia, Brazil; Federal University of Minas Gerais, Belo Horizonte, Brazil
Objective. This study aimed to evaluate HLA-G expression in primary oral cavity squamous cell carcinoma (OCSCC) and potentially malignant lesions and to evaluate its relationship with clinicopathologic parameters. Study Design. HLA-G expression in samples from patients with metastatic and nonmetastatic OCSCC (n ¼ 60), potentially malignant lesions (n ¼ 15), and clinically and histologically normal oral mucosa (n ¼ 10) was characterized by immunohistochemistry. The density of CD8, CD83, and CD68 cells and Ki-67þ and bcl-2þ neoplastic cells were analyzed. Results. HLA-G expression by neoplastic cells was significantly higher in metastatic OCSCC compared with nonmetastatic OCSCC (P ¼ .01). Higher HLA-G expression was observed in OCSCC than in potentially malignant lesions (P ¼ .006). Moreover, patients with lower HLA-G expression exhibited a tendency toward longer survival (22 months) compared with those with higher HLA-G expression (16 months). Conclusions. Our findings suggest that increased HLA-G expression in metastatic OCSCC may represent a tumor escape mechanism, which portends an unfavorable clinical prognosis. (Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:361-368)
Human leukocyte antigen G (HLA-G) is a nonclassical major histocompatibility class Ib antigen that plays a significant physiologic role in maternal-fetal immune tolerance.1,2 HLA-G is characterized by limited polymorphism, restricted tissue expression, and alternative splicing that leads to the production of 7 different isoforms (HLA-G1 to G7).1,2 Moreover, HLA-G exerts an overall negative immune function by inhibiting the activity of natural killer cells (NK cells), cytotoxic T lymphocytes (CTLs), and antigen-presenting cells (APCs), which are the main cells involved in the This study was sponsored by grants from the National Council for Scientific and Technological Development (CNPq) (grants 473324/ 2012-1 and 303886/2012-9) and the Foundation for Research Support in the State of Goiás (FAPEG, grant 201210267001105). a Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. b Department of Biology, State University of Goiás, Morrinhos, Brazil; Department of Medicine, Pontifical Catholic University of Goiás. c Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. d Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. e Laboratory of Cellular Pharmacology and Toxicology, Faculty of Pharmacy, Federal University of Goiás. f Laboratory of Cellular Pharmacology and Toxicology, Faculty of Pharmacy, Federal University of Goiás. g Department of Oral Surgery and Pathology, Dental School, Federal University of Minas Gerais. h Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. Received for publication Jun 16, 2013; returned for revision Nov 26, 2013; accepted for publication Dec 1, 2013. Ó 2014 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2013.12.001
development of an effective cytotoxic antitumor immune response.3,4 This immunoregulatory mechanism occurs through a direct interaction with the inhibitory receptors KIR2DL4 (CD158d), expressed by NK cells and T cells; immunoglobulin-like transcript 2 (ILT2) (CD85j) (LILRB1), expressed on the surface of T cells, NK cells, dendritic cells (DCs), and monocytes; and ILT4 (CD85d) (LILRB2), expressed by the myeloid type APCs (dendritic cells and monocytes).3-7 Thus, HLA-G expression may represent a recent mechanism used by tumor cells to resist or escape host immunosurveillance; this mechanism may also be linked to metastasis.8,9 In light of these inhibitory properties, some researchers have focused on the role of HLA-G in antitumor immunology, notably with respect to its capacity for promoting tumor progression and, consequently, influencing the clinical course of disease.4,8,10 Interestingly, HLA-G expression has been correlated with the clinical stage of disease, metastasis, and survival in some neoplasms, including cervical carcinoma,11
Statement of Clinical Relevance HLA-G may contribute to escape from host immunosurveillance, thereby influencing tumor progression, metastasis, and clinical course of disease. To our knowledge, this study was the first to evaluate the expression of HLA-G in patients with metastatic and nonmetastatic oral cavity squamous cell carcinoma. 361
ORAL AND MAXILLOFACIAL PATHOLOGY 362 Gonçalves et al.
OOOO March 2014
Fig. 1. Schematic representation of the study design. (OCSCC, oral cavity squamous cell carcinoma.)
esophageal squamous cell carcinoma,12 and, more recently, nasopharyngeal carcinoma.13 Additionally, Nuckel et al.14 found that in chronic lymphocytic leukemia, HLA-G expression was a better independent prognostic factor than zeta-associated protein 70 (ZAP70) or CD38 status, which are currently used as prognostic markers. Therefore, recent studies have suggested that HLA-G could be a clinical marker in diagnosis or prediction of clinical outcomes.11,15-17 Regarding primary oral cavity squamous cell carcinoma (OCSCC), to our knowledge there has been only one study that compared HLA-G expression in benign, premalignant, and malignant oral lesions.18 However, the relationship between HLA-G and clinical prognostic factors (especially lymph node metastasis and patient survival) has not yet been defined.18 This study was conducted to evaluate HLA-G expression in neoplastic cells and the OCSCC microen vironment, as well as the relationship between HLA-G expression with lymph node metastasis, age, gender, ethnic group, alcohol and tobacco consumption, tumor location, T stage, clinical outcome, survival data, histologic grading according to the World Health Organization (WHO)19 and Anneroth and Hansen,20 depth of invasion, tumoral proliferation, apoptosis, and density of immune-inflammatory cells.
MATERIALS AND METHODS Samples This study was approved by the Institutional Ethics Committee for Human Subjects (Protocol 032/2011). The samples for this retrospective study were obtained from 60 patients with primary OCSCC, comprising metastatic OCSCC (group 1) (n ¼ 30) and nonmetastatic OCSCC (group 2) (n ¼ 30). We also assessed whether HLA-G expression was maintained by metastatic neoplastic cells in cervical lymph nodes compared with neoplastic cells at the primary sites (group 1). All samples were obtained from files at the Division of Anatomopathology and Cytopathology at Araújo Jorge Hospital, Goiás Cancer Combat Association, Goiânia, Brazil. Furthermore, 15 samples of
potentially malignant lesions and 10 clinically and histologically normal oral samples were obtained from the Oral Disease Center of Goiás State (Oral Pathology Laboratory) of the Federal University of Goiás, Brazil (Figure 1). For inclusion in this study, patients were required to have undergone surgical treatment consisting of cervical lymph node removal and to have had a minimum follow-up of 36 months. The exclusion criteria ruled out patients with squamous cell carcinoma at other sites, patients who did not return after surgical removal of the tumor (no follow-up), and patients who had received radiotherapy, chemotherapy, or other treatment before surgery. Medical records were reviewed to determine age, gender, ethnic group, and tobacco and alcohol consumption (yes or no), as well as follow-up information including survival time and death. Light microscopy Specimens were fixed in 10% buffered formalin (pH 7.4) and paraffin embedded. Microscopic features were assessed by analysis of one 5-mm section of each sample stained with hematoxylin-eosin. All lymph node samples were assessed for the presence of metastasis, and the primary OCSCC sections were graded according to the WHO classification of tumors19 and the histologic malignancy grading system developed by Anneroth and Hansen.20 Immunohistochemistry Immunohistochemistry was performed as described previously by Costa et al.21 For HLA-G, marker of proliferation Ki-67 (MKI67), CD8, CD83, and CD68 retrieval, sections were immersed in citrate buffer (pH 6.0), and for bcl-2 (B-cell lymphoma 2) (BCL2) retrieval, sections were immersed in ethylenediaminetetraacetic acid buffer (pH 9.0), both for 25 minutes. The slides were incubated with the following primary mouse monoclonal antibodies: antihuman HLA-G (clone MEM-G2; Exbio, Prague, Czech Republic; at 1:100), antihuman Ki-67
OOOO Volume 117, Number 3
(clone MM1; Novocastra, Newcastle, UK; at 1:100), antihuman bcl-2 (clone 124; Dako, Carpinteria, CA, USA; at 1:500), antihuman CD8 (clone C8/144B; Dako; at 1:200), antihuman CD83 (clone HB15a; Santa Cruz Biotechnology, Santa Cruz, CA, USA; at 1:50) and antihuman CD68 (clone KP1, Novocastra, at 1:1000). The sections containing anti-HLA-G and CD83 antibodies were incubated with Starr Trek Universal HRP Detection System (Biocare Medical, Concord, CA, USA), whereas those containing anti-Ki-67, bcl-2, CD8, and CD68 antibodies were incubated with an LSAB Kit (K0492; Dako). Trophoblast samples were used as positive controls for HLA-G. Negative controls were obtained by the omission of the primary antibody, which was replaced with 1% phosphate-buffered saline with bovine serum albumin and preimmune mouse serum (X501-1; Dako). Evaluation of stained sections HLA-G was evaluated in neoplastic cells (tumoral parenchyma) and cells present in the stroma near the primary invasion front (peritumoral region), as well as in the intraepithelial region of potentially malignant lesions and clinically and histologically normal oral mucosa. A modified semiquantitative scoring system adopted from previous studies was used.22,23 The percentage of positive tumor cells was scored as: 0, no tumor cells stained; 1, < 25% of cells stained; and 2, 25% of cells stained. Staining intensity was scored as: 0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining. The immunoreactive score (IRS) was calculated by multiplying the percentage of positive cells (scored 0 to 2) by the staining intensity (scored 0 to 3). HLA-G expression in tumors with IRS ¼ 0 was considered absent and for IRS 2 was considered low, whereas tumors with IRS > 2 were considered to have high HLA-G expression. All sections were analyzed in a blinded manner using a light microscope at high magnification (400). The depth of invasion (in millimeters) was assessed by means of a scale linked to the light microscope with high-power fields (100) that allowed us to investigate the tumor depth of invasion from the surface of the epithelium to the deepest tumoral invasion front. Counts of the density (per square millimeter) of immune-inflammatory cells (CD8, CD83, and CD68) and the proportion of neoplastic cells (Ki-67 and bcl-2) were performed in 10 alternate microscopic highpower fields (400) using an integration graticule (4740680000000-Netzmik-rometer 12.5, Carl Zeiss, Göttingen, Germany). At this magnification, each graticule integration field presented an area of 0.0961 mm2. Descriptive analyses were expressed as the means standard deviation of the number of observations per square millimeter.
ORIGINAL ARTICLE Gonçalves et al. 363
Statistical analysis All statistical analyses were performed with the aid of SPSS software (version 17.0; SPSS Inc). Comparative analyses between the groups were performed using the Fisher exact test or Pearson c2 test. To analyze the relationship between HLA-G and microscopic parameters (CTLs CD8þ, CD83þ DCs, and macrophage CD68þ), the Mann-Whitney test was used. Survival time was calculated from surgical resection until the last follow-up appointment or the death of the patient. Survival differences between the groups were evaluated by using the log-rank test. A value of P < .05 was considered significant.
RESULTS The main clinical and histologic data for our group of 60 patients with primary OCSCC (30 with cervical lymph node metastasis and 30 without lymph node metastasis) are summarized in Tables I and II. Analysis of cases of metastatic and nonmetastatic OCSCC found a predominance of males (70% and 60%, respectively) (Table I). Mean survival time was 19.5 months (CI, 079) for patients with metastatic OCSCC and 31.6 months (CI, 4-131) for patients with nonmetastatic OCSCC (Table I). Immunohistochemical staining found that cells expressing HLA-G exhibited cytoplasmic and membrane staining, presenting a staining intensity ranging from weak to strong (Figure 2, A-D). We found that HLA-G was expressed by both neoplastic cells (tumoral parenchyma) and immune-inflammatory cells; lymphocytes (T or B) and plasma cells were diffusely distributed over the tumor stroma (Figure 3). Moreover, it was observed that immuneinflammatory cells tended to lose the expression of HLA-G when distant from the tumor invasion front. In addition, the higher the atypia presented by neoplastic cells, the greater the intensity of expression of HLA-G. Comparative analysis showed significantly higher HLA-G expression (IRS 2) in metastatic OCSCC (72% of samples) than in nonmetastatic OCSCC (28% of samples) (P ¼ .01) (Figure 4, A-B, and Figure 5, A-B). However, there was no significant difference in HLA-G expression between metastatic OCSCC (40% of samples) and nonmetastatic OCSCC (20% of samples) in the peritumoral region (P ¼ .38) (Figure 4, A-B, and Figure 5, A-B). HLA-G expression was significantly higher in primary OCSCC compared with potentially malignant lesions (P ¼ .006). In clinically and histologically normal oral mucosa, 80% presented absence or slight express ion of HLA-Gþ cells (P ¼ .27). The study also determined that 70% of patients with primary OCSCC
ORAL AND MAXILLOFACIAL PATHOLOGY 364 Gonçalves et al.
OOOO March 2014
Table I. Main clinical findings (%) of patients with metastatic OCSCC (n ¼ 30) and nonmetastatic OCSCC (n ¼ 30) Clinical features Age 58 years >58 years Gender Male Female Ethnic group White Nonwhite Alcohol Yes No Tobacco Yes No Location Oral tongue Floor of the mouth T stage T1-T2 T3-T4 Clinical outcome Dead Alive (overall survival) Survival time 36 months 4.7 mm Ki-67 proportion (mean ¼ 17.1%; 1.00 28.70) 17.1 >17.1 bcl-2 proportion (mean ¼ 1.02%; 0.38 2.56) 1.02 >1.02
Metastatic OCSCC Nonmetastatic n (%) OCSCC n (%)
P
18 (60) 12 (40)
18 (60) 12 (40)
.85
18 (60) 12 (40)
12 (40) 18 (60)
.06
21 (70) 9 (30)
18 (60) 12 (40)
.73
27 (90) 3 (10)
21 (70) 9 (30)
.77
3 (10) 27 (90)
6 (20) 27 (80)
.20
6 (20) 24 (80)
9 (30) 21 (70)
.45
18 (60) 12 (40)
12 (40) 18 (60)
.63
9 (30) 21 (70)
18 (60) 12 (40)
.02
13 (43.3) 17 (56.7)
15 (50) 15 (50)
1.0
16 (53.4) 14 (46.6)
12 (40) 18 (60)
1.0
OCSCC, oral cavity squamous cell carcinoma; bcl-2, B-cell lymphoma 2 (BCL2); Ki-67, marker of proliferation Ki-67 (MKI67); WHO, World Health Organization.
clinical outcome (see Table I); the WHO classification of tumors19 and the histologic malignancy grading system developed by Anneroth and Hansen20 (tendency to keratinization, nuclear aberrations, number of mitoses above basal cell layer, mode of invasion, stage of invasion, and inflammatory response); tumoral proliferation (Ki-67þ neoplastic cells); tumoral apoptosis (bcl-2þ neoplastic cells) (see Table II); and density of immune-inflammatory cells (CTLs CD8þ, P ¼ .48; CD83þ DCs, P ¼ .32; and macrophage CD68þ, P ¼ .14) (Table III).
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Gonçalves et al. 365
Fig. 2. Photomicrograph showing different categories of immunolabeling intensity. A, No HLA-G staining in a moderately differentiated nonmetastatic OCSCC of the oral tongue. B, Weak (IRS ¼ 1) staining by neoplastic cells in a moderately differentiated metastatic OCSCC of the oral tongue. C, Strong (IRS ¼ 6) staining by neoplastic cells in a moderately differentiated metastatic OCSCC of the floor of the mouth. D, Trophoblast samples (positive control) were used as a standard for analyzing the intensity of immunolabeling (immunohistochemical staining, original magnification 400) (scale ¼ 50 mm). (IRS, immunoreactive score.)
Fig. 3. Photomicrograph showing immunoexpression of HLA-G by immune-inflammatory cells such as lymphocytes (T or B) and plasma cells (peritumoral region) in a metastatic oral cavity squamous cell carcinoma of the floor of the mouth with marked inflammatory infiltrate (immunohistochemical staining, original magnification 400) (scale ¼ 50 mm).
DISCUSSION Most cancers tend to develop several immunomodulatory strategies that allow them to evade or resist an antitumor response. Of these, HLA-G has been identified as being capable of exerting a significant inhibitory
effect on immune effector cells.3,7,24-27 To our knowledge, the present study is the first to find that HLA-G was significantly augmented in metastatic OCSCC compared with nonmetastatic OCSCC. In addition, there was a tendency toward longer survival for patients with absent or low HLA-G expression compared with those with high HLA-G expression. We speculate that HLA-G expression by neoplastic cells may lead to an escape from host immunosurveillance, thus contributing to tumor progression and metastasis. Consistent with the aforementioned findings, Yie et al.,12 Lin et al.,28 Nunes et al.,29 and Cai et al.,13 in studies with esophageal squamous cell carcinoma (ESCC), papillary thyroid cancer, and nasopharyngeal carcinoma, found that HLA-G expression was significantly correlated with metastasis and reduced survival. These researchers did not identify any significant association between high HLA-G expression and other clinicopathologic characteristics, including tumoral proliferation, apoptosis, and histologic grading. However, Yoon et al.,11 in a study on cervical cancer, reported that HLA-G expression in early-stage patients was significantly greater than in advanced-stage patients. Yie et al.12 observed that in ESCC, HLA-G expression cancer cells was significantly correlated not only with metastasis and clinical stage but also with histologic grade, depth of invasion, and host immune
ORAL AND MAXILLOFACIAL PATHOLOGY 366 Gonçalves et al.
OOOO March 2014
Fig. 4. A, Photomicrograph showing immunoexpression of HLA-G by neoplastic cells in metastatic OCSCC of the floor of the mouth, showing intense staining and a high percentage of HLA-Gþ cells (IRS ¼ 6) (immunohistochemical staining, original magnification 200; scale ¼ 100 mm). B, Photomicrograph showing immunoexpression of HLA-G by neoplastic cells in nonmetastatic OCSCC of the oral tongue with groups of infiltrating cells showing weak staining and a low percentage of HLA-Gþ cells (IRS ¼ 1) (immunohistochemical staining, original magnification 400; scale ¼ 50 mm). (IRS, immunoreactive score.)
Fig. 5. High HLA-G expression (A) in tumoral parenchyma and (B) in the peritumoral region in samples of patients with metastatic and nonmetastatic OCSCC. The Fisher exact test was used to compare groups at a .05 significance level: there was a statistically significant difference when comparing metastatic and nonmetastatic OCSCC in tumoral parenchyma (A); however, this pattern was not observed in the peritumoral region (B) (P ¼ .01 and .38, respectively).
response. Our study also found a statistically significant relationship between higher HLA-G expression and increased depth of invasion. These data allow us to hypothesize that high HLA-G expression by neoplastic cells may lead to the infiltration of such cells because this contributes to an escape from host immunosurveillance. This, along with other factors, may consequently lead to metastasis. However, this study
did not find any correlation between HLA-G and other clinicopathologic parameters. In addition, our findings suggest that differences in the microenvironment of different types of cancer, which may comprise different locations, histologic grade, and clinical stage may also cause them to present different responses to the same immunosuppressive factor. There is only one study comparing HLA-G expression between benign, premalignant, and malignant oral lesions.18 In this study, HLA-G expression in benign and premalignant oral lesions was higher than in malignant tissues. The authors suggested that HLA-G could be associated with the early stages of carcinogenesis but did not exclude the possibility that genetic alterations may modify HLA-G expression in neoplastic cells.18 In contrast to these findings, the current study identified higher levels of HLA-G expression in primary OCSCC than in potentially malignant lesions. Therefore, we believe that the presence of HLA-G in potentially malignant lesions may contribute to the escape of altered cells and consequently to modifications in the microenvironment during the initial stages of tumorigenesis. In the study by Fregonezi et al.,18 correlation between HLA-G expression and clinical prognostic factors was not assessed. The present study introduces new data to the literature and, most importantly, reports an association between HLA-G and metastasis in primary OCSCC. It is well established that HLA-G may inhibit the functions of major immune cells, including DCs, CTLs, and monocytes.3-7,9 Our study found a higher density of CTLs (CD8þ cells) in patients with absent or low HLAG expression compared with those with high HLA-G expression by neoplastic cells; however, this relationship was not statistically significant. Because this study has not evaluated whether the function of the analyzed cells was compromised, we believe that further studies in patients with primary OCSCC are required.
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Gonçalves et al. 367
Fig. 6. A, Photomicrograph from a patient with moderately differentiated primary OCSCC of the floor of the mouth with metastasis to cervical lymph nodes. We have observed high HLA-G expression (immunoreactive score ¼ 6) by neoplastic cells infiltrating in solid cords and sometimes in isolation. B, Photomicrograph showing a metastatic cervical lymph node demonstrating maintenance of HLA-G expression by metastatic neoplastic cells (*) in the lymph nodes of the same patient with primary OCSCC (A) (immunohistochemical staining, original magnification 400; scale ¼ 50 mm).
Table III. Mean and SD of immune-inflammatory cells (CD8þ, CD83þ, and CD68þ) in the peritumoral region in patients with OCSCC with absent or low and high HLA-G expressions Immune-inflammatory cells CD8þ CD83þ CD68þ
Fig. 7. Kaplan-Meier survival curve, according to HLA-G expression in patients with primary oral cavity squamous cell carcinoma. HLA-G expression was dichotomized into 2 groups: absent or low HLA-G expression by neoplastic cells; and high HLA-G expression by neoplastic cells (log-rank; P ¼ .67).
We also found that HLA-Gþ cells in the stroma of primary OCSCC consisted of an inflammatory infiltrate of mononuclear cells, including lymphocytes (T or B) and plasma cells, and that once distant from the tumor invasion front, they tended to lose their HLA-G expression. These results could be associated with another mechanism employed by HLA-Gþ cells termed trogocytosis.4,24 Trogocytosis consists of a rapid process in which a cell acquires the cytoplasmic content or membrane fragments of another cell through cell-to-cell contact.4,30,31 In this context, Wiendl et al.32 found that 10% or more of HLA-Gþ cells are necessary to inhibit lysis of HLA-G cells in human gliomas. Thus, a few cells expressing HLA-G would be sufficient to exert an important inhibitory effect, culminating in greater tumor protection. However, such mechanisms need to be further analyzed. In summary, the findings in this study suggest that increased HLA-G expression in metastatic OCSCC
Cells/mm2 HLA-G expression
Mean
SD
Absent or low HLA-G High HLA-G Absent or low HLA-G High HLA-G Absent or low HLA-G High HLA-G
367.5 201.6 9.2 9.1 154.5 268.3
157.4 100.6 5.6 3.3 103.5 175.6
P > .05 for all groups compared (patients with absent or low HLA-G expression vs patients with high HLA-G expression) (Mann-Whitney test). SD, standard deviation.
may represent a recent escape mechanism used by neoplastic cells, which portends unfavorable clinical outcome considering the scarcity of reports on the function of HLA-G in primary OCSCC, further investigations are necessary to establish the precise role of HLA-G in primary OCSCC progression. The authors thank Araújo Jorge Hospital, Goiás Cancer Combat Association, Goiânia, Brazil, with special appreciation to Dr Rita de Cássia Gonçalves Alencar.
REFERENCES 1. Kovats S, Main EK, Librach C, Stubblebine M, Fisher SJ, DeMars R. A class I antigen, HLA-G, expressed in human trophoblasts. Science. 1990;248:220-223. 2. Rouas-Freiss N, Goncalves RM, Menier C, Dausset J, Carosella ED. Direct evidence to support the role of HLA-G in protecting the fetus from maternal uterine natural killer cytolysis. Proc Natl Acad Sci U S A. 1997;94:11520-11525. 3. Rouas-Freiss N, Moreau P, Ferrone S, Carosella ED. HLA-G proteins in cancer: do they provide tumor cells with an escape mechanism? Cancer Res. 2005;65:10139-10144.
ORAL AND MAXILLOFACIAL PATHOLOGY 368 Gonçalves et al. 4. Rouas-Freiss N, Moreau P, Menier C, LeMaoult J, Carosella ED. Expression of tolerogenic HLA-G molecules in cancer prevents antitumor responses. Semin Cancer Biol. 2007;17:413-421. 5. Mouillot G, Marcou C, Rousseau P, Rouas-Freiss N, Carosella ED, Moreau P. HLA-G gene activation in tumor cells involves cisacting epigenetic changes. Int J Cancer. 2005;113:928-936. 6. Moreau P, Mouillot G, Rousseau P, Marcou C, Dausset J, Carosella ED. HLA-G gene repression is reversed by demethylation. Proc Natl Acad Sci U S A. 2003;100:1191-1196. 7. Bukur J, Rebmann V, Grosse-Wilde H, et al. Functional role of human leukocyte antigen-G up-regulation in renal cell carcinoma. Cancer Res. 2003;63:4107-4111. 8. Paul P, Rouas-Freiss N, Khalil-Daher I, et al. HLA-G expression in melanoma: a way for tumor cells to escape from immunosurveillance. Proc Natl Acad Sci U S A. 1998;95:4510-4515. 9. Zidi I, Ben Amor N. HLA-G as predisposing for metastasis. Med Hypotheses. 2011;77:134-139. 10. Cabestre FA, Lefebvre S, Moreau P, et al. HLA-G expression: immune privilege for tumour cells? Semin Cancer Biol. 1999;9:27-36. 11. Yoon BS, Kim YT, Kim JW, Kim SH, Kim JH, Kim SW. Expression of human leukocyte antigen-G and its correlation with interleukin-10 expression in cervical carcinoma. Int J Gynaecol Obstet. 2007;98:48-53. 12. Yie SM, Yang H, Ye SR, Li K, Dong DD, Lin XM. Expression of HLA-G is associated with prognosis in esophageal squamous cell carcinoma. Am J Clin Pathol. 2007;128:1002-1009. 13. Cai MB, Han HQ, Bei JX, et al. Expression of human leukocyte antigen G is associated with prognosis in nasopharyngeal carcinoma. Int J Biol Sci. 2012;8:891-900. 14. Nuckel H, Rebmann V, Durig J, Duhrsen U, Grosse-Wilde H. HLA-G expression is associated with an unfavorable outcome and immunodeficiency in chronic lymphocytic leukemia. Blood. 2005;105:1694-1698. 15. Ibrahim EC, Aractingi S, Allory Y, et al. Analysis of HLA antigen expression in benign and malignant melanocytic lesions reveals that upregulation of HLA-G expression correlates with malignant transformation, high inflammatory infiltration and HLA-A1 genotype. Int J Cancer. 2004;108:243-250. 16. Davidson B, Elstrand MB, McMaster MT, et al. HLA-G expression in effusions is a possible marker of tumor susceptibility to chemotherapy in ovarian carcinoma. Gynecol Oncol. 2005;96:42-47. 17. Alkhouly N, Shehata I, Ahmed MB, Shehata H, Hassan S, Ibrahim T. HLA-G expression in acute lymphoblastic leukemia: a significant prognostic tumor biomarker. Med Oncol. 2013;30:460. 18. Fregonezi PA, Silva TG, Simoes RT, et al. Expression of nonclassical molecule human leukocyte antigen-G in oral lesions. Am J Otolaryngol. 2012;33:193-198. 19. Pindborg JJ, Reichart PA, Smith CJ, van der Wall I. World Health Organization International Histological Classification of Tumours: Histological Typing of Cancer and Precancer of the Oral Mucosa. Berlin, Germany: Springer-Verlag; 1997: 21-31.
OOOO March 2014 20. Anneroth G, Hansen LS. A methodologic study of histologic classification and grading of malignancy in oral squamous cell carcinoma. Scand J Dent Res. 1984;92:448-468. 21. Costa NL, Alencar R de C, Valadares MC, Silva TA, Mendonça EF, Batista AC. The clinicopathological significance of the expression of Granzyme B in oral squamous cell carcinoma. Oral Oncol. 2010;46:185-189. 22. Gori S, Sidoni A, Colozza M, et al. EGFR, pMAPK, pAkt and PTEN status by immunohistochemistry: correlation with clinical outcome in HER2-positive metastatic breast cancer patients treated with trastuzumab. Ann Oncol. 2009;20:648-654. 23. Hiraishi Y, Wada T, Nakatani K, Negoro K, Fujita S. Immunohistochemical expression of EGFR and p-EGFR in oral squamous cell carcinomas. Pathol Oncol Res. 2006;12:87-91. 24. Curigliano G, Criscitiello C, Gelao L, Goldhirsch A. Molecular pathways: human leukocyte antigen G (HLA-G). Clin Cancer Res. 2013;19:5564-5571. 25. Carosella ED, HoWangYin KY, Favier B, LeMaoult J. HLAG-dependent suppressor cells: diverse by nature, function, and significance. Hum Immunol. 2008;69:700-707. 26. Rouas-Freiss N, Moreau P, Menier C, Carosella ED. HLA-G in cancer: a way to turn off the immune system. Semin Cancer Biol. 2003;13:325-336. 27. Seliger B, Abken H, Ferrone S. HLA-G and MIC expression in tumors and their role in anti-tumor immunity. Trends Immunol. 2003;24:82-87. 28. Lin A, Zhang X, Zhou WJ, et al. Human leukocyte antigen-G expression is associated with a poor prognosis in patients with esophageal squamous cell carcinoma. Int J Cancer. 2011;129: 1382-1390. 29. Nunes LM, Ayres FM, Francescantonio IC, et al. Association between the HLA-G molecule and lymph node metastasis in papillary thyroid cancer. Hum Immunol. 2013;74:447-451. 30. Carosella ED, Moreau P, Lemaoult J, Rouas-Freiss N. HLA-G: from biology to clinical benefits. Trends Immunol. 2008;29: 125-132. 31. Wastowski IJ, Sampaio-Barros PD, Amstalden EM, et al. HLA-G expression in the skin of patients with systemic sclerosis. J Rheumatol. 2009;36:1230-1234. 32. Wiendl H, Mitsdoerffer M, Hofmeister V, et al. A functional role of HLA-G expression in human gliomas: an alternative strategy of immune escape. J Immunol. 2002;168:4772-4780. Reprint requests: Aline Carvalho Batista Disciplina de Patologia Geral e Bucal Faculdade de Odontologia Universidade Federal de Goiás Praça Universitária S/N Setor Universitário CEP 74605-220 Brazil [email protected]
The clinicopathologic significance of the expression of HLA-G in oral squamous cell carcinoma Andréia Souza Gonçalves, DDS,a Isabela Jubé Wastowski, BS, PhD,b Lucas Raineri Capeletti,c Nancy Tomoko Sacono, DDS, PhD,d Alane Pereira Cortez, BS, PhD,e Marize Campos Valadares, BS, PhD,f Tarcília Aparecida Silva, DDS, PhD,g and Aline Carvalho Batista, DDS, PhDh Federal University of Goiás, Goiânia, Brazil; State University of Goiás, Morrinhos, Brazil; Pontifical Catholic University of Goiás, Goiânia, Brazil; Federal University of Minas Gerais, Belo Horizonte, Brazil
Objective. This study aimed to evaluate HLA-G expression in primary oral cavity squamous cell carcinoma (OCSCC) and potentially malignant lesions and to evaluate its relationship with clinicopathologic parameters. Study Design. HLA-G expression in samples from patients with metastatic and nonmetastatic OCSCC (n ¼ 60), potentially malignant lesions (n ¼ 15), and clinically and histologically normal oral mucosa (n ¼ 10) was characterized by immunohistochemistry. The density of CD8, CD83, and CD68 cells and Ki-67þ and bcl-2þ neoplastic cells were analyzed. Results. HLA-G expression by neoplastic cells was significantly higher in metastatic OCSCC compared with nonmetastatic OCSCC (P ¼ .01). Higher HLA-G expression was observed in OCSCC than in potentially malignant lesions (P ¼ .006). Moreover, patients with lower HLA-G expression exhibited a tendency toward longer survival (22 months) compared with those with higher HLA-G expression (16 months). Conclusions. Our findings suggest that increased HLA-G expression in metastatic OCSCC may represent a tumor escape mechanism, which portends an unfavorable clinical prognosis. (Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:361-368)
Human leukocyte antigen G (HLA-G) is a nonclassical major histocompatibility class Ib antigen that plays a significant physiologic role in maternal-fetal immune tolerance.1,2 HLA-G is characterized by limited polymorphism, restricted tissue expression, and alternative splicing that leads to the production of 7 different isoforms (HLA-G1 to G7).1,2 Moreover, HLA-G exerts an overall negative immune function by inhibiting the activity of natural killer cells (NK cells), cytotoxic T lymphocytes (CTLs), and antigen-presenting cells (APCs), which are the main cells involved in the This study was sponsored by grants from the National Council for Scientific and Technological Development (CNPq) (grants 473324/ 2012-1 and 303886/2012-9) and the Foundation for Research Support in the State of Goiás (FAPEG, grant 201210267001105). a Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. b Department of Biology, State University of Goiás, Morrinhos, Brazil; Department of Medicine, Pontifical Catholic University of Goiás. c Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. d Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. e Laboratory of Cellular Pharmacology and Toxicology, Faculty of Pharmacy, Federal University of Goiás. f Laboratory of Cellular Pharmacology and Toxicology, Faculty of Pharmacy, Federal University of Goiás. g Department of Oral Surgery and Pathology, Dental School, Federal University of Minas Gerais. h Department of Stomatology (Oral Pathology), Dental School, Federal University of Goiás. Received for publication Jun 16, 2013; returned for revision Nov 26, 2013; accepted for publication Dec 1, 2013. Ó 2014 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2013.12.001
development of an effective cytotoxic antitumor immune response.3,4 This immunoregulatory mechanism occurs through a direct interaction with the inhibitory receptors KIR2DL4 (CD158d), expressed by NK cells and T cells; immunoglobulin-like transcript 2 (ILT2) (CD85j) (LILRB1), expressed on the surface of T cells, NK cells, dendritic cells (DCs), and monocytes; and ILT4 (CD85d) (LILRB2), expressed by the myeloid type APCs (dendritic cells and monocytes).3-7 Thus, HLA-G expression may represent a recent mechanism used by tumor cells to resist or escape host immunosurveillance; this mechanism may also be linked to metastasis.8,9 In light of these inhibitory properties, some researchers have focused on the role of HLA-G in antitumor immunology, notably with respect to its capacity for promoting tumor progression and, consequently, influencing the clinical course of disease.4,8,10 Interestingly, HLA-G expression has been correlated with the clinical stage of disease, metastasis, and survival in some neoplasms, including cervical carcinoma,11
Statement of Clinical Relevance HLA-G may contribute to escape from host immunosurveillance, thereby influencing tumor progression, metastasis, and clinical course of disease. To our knowledge, this study was the first to evaluate the expression of HLA-G in patients with metastatic and nonmetastatic oral cavity squamous cell carcinoma. 361
ORAL AND MAXILLOFACIAL PATHOLOGY 362 Gonçalves et al.
OOOO March 2014
Fig. 1. Schematic representation of the study design. (OCSCC, oral cavity squamous cell carcinoma.)
esophageal squamous cell carcinoma,12 and, more recently, nasopharyngeal carcinoma.13 Additionally, Nuckel et al.14 found that in chronic lymphocytic leukemia, HLA-G expression was a better independent prognostic factor than zeta-associated protein 70 (ZAP70) or CD38 status, which are currently used as prognostic markers. Therefore, recent studies have suggested that HLA-G could be a clinical marker in diagnosis or prediction of clinical outcomes.11,15-17 Regarding primary oral cavity squamous cell carcinoma (OCSCC), to our knowledge there has been only one study that compared HLA-G expression in benign, premalignant, and malignant oral lesions.18 However, the relationship between HLA-G and clinical prognostic factors (especially lymph node metastasis and patient survival) has not yet been defined.18 This study was conducted to evaluate HLA-G expression in neoplastic cells and the OCSCC microen vironment, as well as the relationship between HLA-G expression with lymph node metastasis, age, gender, ethnic group, alcohol and tobacco consumption, tumor location, T stage, clinical outcome, survival data, histologic grading according to the World Health Organization (WHO)19 and Anneroth and Hansen,20 depth of invasion, tumoral proliferation, apoptosis, and density of immune-inflammatory cells.
MATERIALS AND METHODS Samples This study was approved by the Institutional Ethics Committee for Human Subjects (Protocol 032/2011). The samples for this retrospective study were obtained from 60 patients with primary OCSCC, comprising metastatic OCSCC (group 1) (n ¼ 30) and nonmetastatic OCSCC (group 2) (n ¼ 30). We also assessed whether HLA-G expression was maintained by metastatic neoplastic cells in cervical lymph nodes compared with neoplastic cells at the primary sites (group 1). All samples were obtained from files at the Division of Anatomopathology and Cytopathology at Araújo Jorge Hospital, Goiás Cancer Combat Association, Goiânia, Brazil. Furthermore, 15 samples of
potentially malignant lesions and 10 clinically and histologically normal oral samples were obtained from the Oral Disease Center of Goiás State (Oral Pathology Laboratory) of the Federal University of Goiás, Brazil (Figure 1). For inclusion in this study, patients were required to have undergone surgical treatment consisting of cervical lymph node removal and to have had a minimum follow-up of 36 months. The exclusion criteria ruled out patients with squamous cell carcinoma at other sites, patients who did not return after surgical removal of the tumor (no follow-up), and patients who had received radiotherapy, chemotherapy, or other treatment before surgery. Medical records were reviewed to determine age, gender, ethnic group, and tobacco and alcohol consumption (yes or no), as well as follow-up information including survival time and death. Light microscopy Specimens were fixed in 10% buffered formalin (pH 7.4) and paraffin embedded. Microscopic features were assessed by analysis of one 5-mm section of each sample stained with hematoxylin-eosin. All lymph node samples were assessed for the presence of metastasis, and the primary OCSCC sections were graded according to the WHO classification of tumors19 and the histologic malignancy grading system developed by Anneroth and Hansen.20 Immunohistochemistry Immunohistochemistry was performed as described previously by Costa et al.21 For HLA-G, marker of proliferation Ki-67 (MKI67), CD8, CD83, and CD68 retrieval, sections were immersed in citrate buffer (pH 6.0), and for bcl-2 (B-cell lymphoma 2) (BCL2) retrieval, sections were immersed in ethylenediaminetetraacetic acid buffer (pH 9.0), both for 25 minutes. The slides were incubated with the following primary mouse monoclonal antibodies: antihuman HLA-G (clone MEM-G2; Exbio, Prague, Czech Republic; at 1:100), antihuman Ki-67
OOOO Volume 117, Number 3
(clone MM1; Novocastra, Newcastle, UK; at 1:100), antihuman bcl-2 (clone 124; Dako, Carpinteria, CA, USA; at 1:500), antihuman CD8 (clone C8/144B; Dako; at 1:200), antihuman CD83 (clone HB15a; Santa Cruz Biotechnology, Santa Cruz, CA, USA; at 1:50) and antihuman CD68 (clone KP1, Novocastra, at 1:1000). The sections containing anti-HLA-G and CD83 antibodies were incubated with Starr Trek Universal HRP Detection System (Biocare Medical, Concord, CA, USA), whereas those containing anti-Ki-67, bcl-2, CD8, and CD68 antibodies were incubated with an LSAB Kit (K0492; Dako). Trophoblast samples were used as positive controls for HLA-G. Negative controls were obtained by the omission of the primary antibody, which was replaced with 1% phosphate-buffered saline with bovine serum albumin and preimmune mouse serum (X501-1; Dako). Evaluation of stained sections HLA-G was evaluated in neoplastic cells (tumoral parenchyma) and cells present in the stroma near the primary invasion front (peritumoral region), as well as in the intraepithelial region of potentially malignant lesions and clinically and histologically normal oral mucosa. A modified semiquantitative scoring system adopted from previous studies was used.22,23 The percentage of positive tumor cells was scored as: 0, no tumor cells stained; 1, < 25% of cells stained; and 2, 25% of cells stained. Staining intensity was scored as: 0, no staining; 1, weak staining; 2, moderate staining; and 3, strong staining. The immunoreactive score (IRS) was calculated by multiplying the percentage of positive cells (scored 0 to 2) by the staining intensity (scored 0 to 3). HLA-G expression in tumors with IRS ¼ 0 was considered absent and for IRS 2 was considered low, whereas tumors with IRS > 2 were considered to have high HLA-G expression. All sections were analyzed in a blinded manner using a light microscope at high magnification (400). The depth of invasion (in millimeters) was assessed by means of a scale linked to the light microscope with high-power fields (100) that allowed us to investigate the tumor depth of invasion from the surface of the epithelium to the deepest tumoral invasion front. Counts of the density (per square millimeter) of immune-inflammatory cells (CD8, CD83, and CD68) and the proportion of neoplastic cells (Ki-67 and bcl-2) were performed in 10 alternate microscopic highpower fields (400) using an integration graticule (4740680000000-Netzmik-rometer 12.5, Carl Zeiss, Göttingen, Germany). At this magnification, each graticule integration field presented an area of 0.0961 mm2. Descriptive analyses were expressed as the means standard deviation of the number of observations per square millimeter.
ORIGINAL ARTICLE Gonçalves et al. 363
Statistical analysis All statistical analyses were performed with the aid of SPSS software (version 17.0; SPSS Inc). Comparative analyses between the groups were performed using the Fisher exact test or Pearson c2 test. To analyze the relationship between HLA-G and microscopic parameters (CTLs CD8þ, CD83þ DCs, and macrophage CD68þ), the Mann-Whitney test was used. Survival time was calculated from surgical resection until the last follow-up appointment or the death of the patient. Survival differences between the groups were evaluated by using the log-rank test. A value of P < .05 was considered significant.
RESULTS The main clinical and histologic data for our group of 60 patients with primary OCSCC (30 with cervical lymph node metastasis and 30 without lymph node metastasis) are summarized in Tables I and II. Analysis of cases of metastatic and nonmetastatic OCSCC found a predominance of males (70% and 60%, respectively) (Table I). Mean survival time was 19.5 months (CI, 079) for patients with metastatic OCSCC and 31.6 months (CI, 4-131) for patients with nonmetastatic OCSCC (Table I). Immunohistochemical staining found that cells expressing HLA-G exhibited cytoplasmic and membrane staining, presenting a staining intensity ranging from weak to strong (Figure 2, A-D). We found that HLA-G was expressed by both neoplastic cells (tumoral parenchyma) and immune-inflammatory cells; lymphocytes (T or B) and plasma cells were diffusely distributed over the tumor stroma (Figure 3). Moreover, it was observed that immuneinflammatory cells tended to lose the expression of HLA-G when distant from the tumor invasion front. In addition, the higher the atypia presented by neoplastic cells, the greater the intensity of expression of HLA-G. Comparative analysis showed significantly higher HLA-G expression (IRS 2) in metastatic OCSCC (72% of samples) than in nonmetastatic OCSCC (28% of samples) (P ¼ .01) (Figure 4, A-B, and Figure 5, A-B). However, there was no significant difference in HLA-G expression between metastatic OCSCC (40% of samples) and nonmetastatic OCSCC (20% of samples) in the peritumoral region (P ¼ .38) (Figure 4, A-B, and Figure 5, A-B). HLA-G expression was significantly higher in primary OCSCC compared with potentially malignant lesions (P ¼ .006). In clinically and histologically normal oral mucosa, 80% presented absence or slight express ion of HLA-Gþ cells (P ¼ .27). The study also determined that 70% of patients with primary OCSCC
ORAL AND MAXILLOFACIAL PATHOLOGY 364 Gonçalves et al.
OOOO March 2014
Table I. Main clinical findings (%) of patients with metastatic OCSCC (n ¼ 30) and nonmetastatic OCSCC (n ¼ 30) Clinical features Age 58 years >58 years Gender Male Female Ethnic group White Nonwhite Alcohol Yes No Tobacco Yes No Location Oral tongue Floor of the mouth T stage T1-T2 T3-T4 Clinical outcome Dead Alive (overall survival) Survival time 36 months 4.7 mm Ki-67 proportion (mean ¼ 17.1%; 1.00 28.70) 17.1 >17.1 bcl-2 proportion (mean ¼ 1.02%; 0.38 2.56) 1.02 >1.02
Metastatic OCSCC Nonmetastatic n (%) OCSCC n (%)
P
18 (60) 12 (40)
18 (60) 12 (40)
.85
18 (60) 12 (40)
12 (40) 18 (60)
.06
21 (70) 9 (30)
18 (60) 12 (40)
.73
27 (90) 3 (10)
21 (70) 9 (30)
.77
3 (10) 27 (90)
6 (20) 27 (80)
.20
6 (20) 24 (80)
9 (30) 21 (70)
.45
18 (60) 12 (40)
12 (40) 18 (60)
.63
9 (30) 21 (70)
18 (60) 12 (40)
.02
13 (43.3) 17 (56.7)
15 (50) 15 (50)
1.0
16 (53.4) 14 (46.6)
12 (40) 18 (60)
1.0
OCSCC, oral cavity squamous cell carcinoma; bcl-2, B-cell lymphoma 2 (BCL2); Ki-67, marker of proliferation Ki-67 (MKI67); WHO, World Health Organization.
clinical outcome (see Table I); the WHO classification of tumors19 and the histologic malignancy grading system developed by Anneroth and Hansen20 (tendency to keratinization, nuclear aberrations, number of mitoses above basal cell layer, mode of invasion, stage of invasion, and inflammatory response); tumoral proliferation (Ki-67þ neoplastic cells); tumoral apoptosis (bcl-2þ neoplastic cells) (see Table II); and density of immune-inflammatory cells (CTLs CD8þ, P ¼ .48; CD83þ DCs, P ¼ .32; and macrophage CD68þ, P ¼ .14) (Table III).
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Gonçalves et al. 365
Fig. 2. Photomicrograph showing different categories of immunolabeling intensity. A, No HLA-G staining in a moderately differentiated nonmetastatic OCSCC of the oral tongue. B, Weak (IRS ¼ 1) staining by neoplastic cells in a moderately differentiated metastatic OCSCC of the oral tongue. C, Strong (IRS ¼ 6) staining by neoplastic cells in a moderately differentiated metastatic OCSCC of the floor of the mouth. D, Trophoblast samples (positive control) were used as a standard for analyzing the intensity of immunolabeling (immunohistochemical staining, original magnification 400) (scale ¼ 50 mm). (IRS, immunoreactive score.)
Fig. 3. Photomicrograph showing immunoexpression of HLA-G by immune-inflammatory cells such as lymphocytes (T or B) and plasma cells (peritumoral region) in a metastatic oral cavity squamous cell carcinoma of the floor of the mouth with marked inflammatory infiltrate (immunohistochemical staining, original magnification 400) (scale ¼ 50 mm).
DISCUSSION Most cancers tend to develop several immunomodulatory strategies that allow them to evade or resist an antitumor response. Of these, HLA-G has been identified as being capable of exerting a significant inhibitory
effect on immune effector cells.3,7,24-27 To our knowledge, the present study is the first to find that HLA-G was significantly augmented in metastatic OCSCC compared with nonmetastatic OCSCC. In addition, there was a tendency toward longer survival for patients with absent or low HLA-G expression compared with those with high HLA-G expression. We speculate that HLA-G expression by neoplastic cells may lead to an escape from host immunosurveillance, thus contributing to tumor progression and metastasis. Consistent with the aforementioned findings, Yie et al.,12 Lin et al.,28 Nunes et al.,29 and Cai et al.,13 in studies with esophageal squamous cell carcinoma (ESCC), papillary thyroid cancer, and nasopharyngeal carcinoma, found that HLA-G expression was significantly correlated with metastasis and reduced survival. These researchers did not identify any significant association between high HLA-G expression and other clinicopathologic characteristics, including tumoral proliferation, apoptosis, and histologic grading. However, Yoon et al.,11 in a study on cervical cancer, reported that HLA-G expression in early-stage patients was significantly greater than in advanced-stage patients. Yie et al.12 observed that in ESCC, HLA-G expression cancer cells was significantly correlated not only with metastasis and clinical stage but also with histologic grade, depth of invasion, and host immune
ORAL AND MAXILLOFACIAL PATHOLOGY 366 Gonçalves et al.
OOOO March 2014
Fig. 4. A, Photomicrograph showing immunoexpression of HLA-G by neoplastic cells in metastatic OCSCC of the floor of the mouth, showing intense staining and a high percentage of HLA-Gþ cells (IRS ¼ 6) (immunohistochemical staining, original magnification 200; scale ¼ 100 mm). B, Photomicrograph showing immunoexpression of HLA-G by neoplastic cells in nonmetastatic OCSCC of the oral tongue with groups of infiltrating cells showing weak staining and a low percentage of HLA-Gþ cells (IRS ¼ 1) (immunohistochemical staining, original magnification 400; scale ¼ 50 mm). (IRS, immunoreactive score.)
Fig. 5. High HLA-G expression (A) in tumoral parenchyma and (B) in the peritumoral region in samples of patients with metastatic and nonmetastatic OCSCC. The Fisher exact test was used to compare groups at a .05 significance level: there was a statistically significant difference when comparing metastatic and nonmetastatic OCSCC in tumoral parenchyma (A); however, this pattern was not observed in the peritumoral region (B) (P ¼ .01 and .38, respectively).
response. Our study also found a statistically significant relationship between higher HLA-G expression and increased depth of invasion. These data allow us to hypothesize that high HLA-G expression by neoplastic cells may lead to the infiltration of such cells because this contributes to an escape from host immunosurveillance. This, along with other factors, may consequently lead to metastasis. However, this study
did not find any correlation between HLA-G and other clinicopathologic parameters. In addition, our findings suggest that differences in the microenvironment of different types of cancer, which may comprise different locations, histologic grade, and clinical stage may also cause them to present different responses to the same immunosuppressive factor. There is only one study comparing HLA-G expression between benign, premalignant, and malignant oral lesions.18 In this study, HLA-G expression in benign and premalignant oral lesions was higher than in malignant tissues. The authors suggested that HLA-G could be associated with the early stages of carcinogenesis but did not exclude the possibility that genetic alterations may modify HLA-G expression in neoplastic cells.18 In contrast to these findings, the current study identified higher levels of HLA-G expression in primary OCSCC than in potentially malignant lesions. Therefore, we believe that the presence of HLA-G in potentially malignant lesions may contribute to the escape of altered cells and consequently to modifications in the microenvironment during the initial stages of tumorigenesis. In the study by Fregonezi et al.,18 correlation between HLA-G expression and clinical prognostic factors was not assessed. The present study introduces new data to the literature and, most importantly, reports an association between HLA-G and metastasis in primary OCSCC. It is well established that HLA-G may inhibit the functions of major immune cells, including DCs, CTLs, and monocytes.3-7,9 Our study found a higher density of CTLs (CD8þ cells) in patients with absent or low HLAG expression compared with those with high HLA-G expression by neoplastic cells; however, this relationship was not statistically significant. Because this study has not evaluated whether the function of the analyzed cells was compromised, we believe that further studies in patients with primary OCSCC are required.
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Gonçalves et al. 367
Fig. 6. A, Photomicrograph from a patient with moderately differentiated primary OCSCC of the floor of the mouth with metastasis to cervical lymph nodes. We have observed high HLA-G expression (immunoreactive score ¼ 6) by neoplastic cells infiltrating in solid cords and sometimes in isolation. B, Photomicrograph showing a metastatic cervical lymph node demonstrating maintenance of HLA-G expression by metastatic neoplastic cells (*) in the lymph nodes of the same patient with primary OCSCC (A) (immunohistochemical staining, original magnification 400; scale ¼ 50 mm).
Table III. Mean and SD of immune-inflammatory cells (CD8þ, CD83þ, and CD68þ) in the peritumoral region in patients with OCSCC with absent or low and high HLA-G expressions Immune-inflammatory cells CD8þ CD83þ CD68þ
Fig. 7. Kaplan-Meier survival curve, according to HLA-G expression in patients with primary oral cavity squamous cell carcinoma. HLA-G expression was dichotomized into 2 groups: absent or low HLA-G expression by neoplastic cells; and high HLA-G expression by neoplastic cells (log-rank; P ¼ .67).
We also found that HLA-Gþ cells in the stroma of primary OCSCC consisted of an inflammatory infiltrate of mononuclear cells, including lymphocytes (T or B) and plasma cells, and that once distant from the tumor invasion front, they tended to lose their HLA-G expression. These results could be associated with another mechanism employed by HLA-Gþ cells termed trogocytosis.4,24 Trogocytosis consists of a rapid process in which a cell acquires the cytoplasmic content or membrane fragments of another cell through cell-to-cell contact.4,30,31 In this context, Wiendl et al.32 found that 10% or more of HLA-Gþ cells are necessary to inhibit lysis of HLA-G cells in human gliomas. Thus, a few cells expressing HLA-G would be sufficient to exert an important inhibitory effect, culminating in greater tumor protection. However, such mechanisms need to be further analyzed. In summary, the findings in this study suggest that increased HLA-G expression in metastatic OCSCC
Cells/mm2 HLA-G expression
Mean
SD
Absent or low HLA-G High HLA-G Absent or low HLA-G High HLA-G Absent or low HLA-G High HLA-G
367.5 201.6 9.2 9.1 154.5 268.3
157.4 100.6 5.6 3.3 103.5 175.6
P > .05 for all groups compared (patients with absent or low HLA-G expression vs patients with high HLA-G expression) (Mann-Whitney test). SD, standard deviation.
may represent a recent escape mechanism used by neoplastic cells, which portends unfavorable clinical outcome considering the scarcity of reports on the function of HLA-G in primary OCSCC, further investigations are necessary to establish the precise role of HLA-G in primary OCSCC progression. The authors thank Araújo Jorge Hospital, Goiás Cancer Combat Association, Goiânia, Brazil, with special appreciation to Dr Rita de Cássia Gonçalves Alencar.
REFERENCES 1. Kovats S, Main EK, Librach C, Stubblebine M, Fisher SJ, DeMars R. A class I antigen, HLA-G, expressed in human trophoblasts. Science. 1990;248:220-223. 2. Rouas-Freiss N, Goncalves RM, Menier C, Dausset J, Carosella ED. Direct evidence to support the role of HLA-G in protecting the fetus from maternal uterine natural killer cytolysis. Proc Natl Acad Sci U S A. 1997;94:11520-11525. 3. Rouas-Freiss N, Moreau P, Ferrone S, Carosella ED. HLA-G proteins in cancer: do they provide tumor cells with an escape mechanism? Cancer Res. 2005;65:10139-10144.
ORAL AND MAXILLOFACIAL PATHOLOGY 368 Gonçalves et al. 4. Rouas-Freiss N, Moreau P, Menier C, LeMaoult J, Carosella ED. Expression of tolerogenic HLA-G molecules in cancer prevents antitumor responses. Semin Cancer Biol. 2007;17:413-421. 5. Mouillot G, Marcou C, Rousseau P, Rouas-Freiss N, Carosella ED, Moreau P. HLA-G gene activation in tumor cells involves cisacting epigenetic changes. Int J Cancer. 2005;113:928-936. 6. Moreau P, Mouillot G, Rousseau P, Marcou C, Dausset J, Carosella ED. HLA-G gene repression is reversed by demethylation. Proc Natl Acad Sci U S A. 2003;100:1191-1196. 7. Bukur J, Rebmann V, Grosse-Wilde H, et al. Functional role of human leukocyte antigen-G up-regulation in renal cell carcinoma. Cancer Res. 2003;63:4107-4111. 8. Paul P, Rouas-Freiss N, Khalil-Daher I, et al. HLA-G expression in melanoma: a way for tumor cells to escape from immunosurveillance. Proc Natl Acad Sci U S A. 1998;95:4510-4515. 9. Zidi I, Ben Amor N. HLA-G as predisposing for metastasis. Med Hypotheses. 2011;77:134-139. 10. Cabestre FA, Lefebvre S, Moreau P, et al. HLA-G expression: immune privilege for tumour cells? Semin Cancer Biol. 1999;9:27-36. 11. Yoon BS, Kim YT, Kim JW, Kim SH, Kim JH, Kim SW. Expression of human leukocyte antigen-G and its correlation with interleukin-10 expression in cervical carcinoma. Int J Gynaecol Obstet. 2007;98:48-53. 12. Yie SM, Yang H, Ye SR, Li K, Dong DD, Lin XM. Expression of HLA-G is associated with prognosis in esophageal squamous cell carcinoma. Am J Clin Pathol. 2007;128:1002-1009. 13. Cai MB, Han HQ, Bei JX, et al. Expression of human leukocyte antigen G is associated with prognosis in nasopharyngeal carcinoma. Int J Biol Sci. 2012;8:891-900. 14. Nuckel H, Rebmann V, Durig J, Duhrsen U, Grosse-Wilde H. HLA-G expression is associated with an unfavorable outcome and immunodeficiency in chronic lymphocytic leukemia. Blood. 2005;105:1694-1698. 15. Ibrahim EC, Aractingi S, Allory Y, et al. Analysis of HLA antigen expression in benign and malignant melanocytic lesions reveals that upregulation of HLA-G expression correlates with malignant transformation, high inflammatory infiltration and HLA-A1 genotype. Int J Cancer. 2004;108:243-250. 16. Davidson B, Elstrand MB, McMaster MT, et al. HLA-G expression in effusions is a possible marker of tumor susceptibility to chemotherapy in ovarian carcinoma. Gynecol Oncol. 2005;96:42-47. 17. Alkhouly N, Shehata I, Ahmed MB, Shehata H, Hassan S, Ibrahim T. HLA-G expression in acute lymphoblastic leukemia: a significant prognostic tumor biomarker. Med Oncol. 2013;30:460. 18. Fregonezi PA, Silva TG, Simoes RT, et al. Expression of nonclassical molecule human leukocyte antigen-G in oral lesions. Am J Otolaryngol. 2012;33:193-198. 19. Pindborg JJ, Reichart PA, Smith CJ, van der Wall I. World Health Organization International Histological Classification of Tumours: Histological Typing of Cancer and Precancer of the Oral Mucosa. Berlin, Germany: Springer-Verlag; 1997: 21-31.
OOOO March 2014 20. Anneroth G, Hansen LS. A methodologic study of histologic classification and grading of malignancy in oral squamous cell carcinoma. Scand J Dent Res. 1984;92:448-468. 21. Costa NL, Alencar R de C, Valadares MC, Silva TA, Mendonça EF, Batista AC. The clinicopathological significance of the expression of Granzyme B in oral squamous cell carcinoma. Oral Oncol. 2010;46:185-189. 22. Gori S, Sidoni A, Colozza M, et al. EGFR, pMAPK, pAkt and PTEN status by immunohistochemistry: correlation with clinical outcome in HER2-positive metastatic breast cancer patients treated with trastuzumab. Ann Oncol. 2009;20:648-654. 23. Hiraishi Y, Wada T, Nakatani K, Negoro K, Fujita S. Immunohistochemical expression of EGFR and p-EGFR in oral squamous cell carcinomas. Pathol Oncol Res. 2006;12:87-91. 24. Curigliano G, Criscitiello C, Gelao L, Goldhirsch A. Molecular pathways: human leukocyte antigen G (HLA-G). Clin Cancer Res. 2013;19:5564-5571. 25. Carosella ED, HoWangYin KY, Favier B, LeMaoult J. HLAG-dependent suppressor cells: diverse by nature, function, and significance. Hum Immunol. 2008;69:700-707. 26. Rouas-Freiss N, Moreau P, Menier C, Carosella ED. HLA-G in cancer: a way to turn off the immune system. Semin Cancer Biol. 2003;13:325-336. 27. Seliger B, Abken H, Ferrone S. HLA-G and MIC expression in tumors and their role in anti-tumor immunity. Trends Immunol. 2003;24:82-87. 28. Lin A, Zhang X, Zhou WJ, et al. Human leukocyte antigen-G expression is associated with a poor prognosis in patients with esophageal squamous cell carcinoma. Int J Cancer. 2011;129: 1382-1390. 29. Nunes LM, Ayres FM, Francescantonio IC, et al. Association between the HLA-G molecule and lymph node metastasis in papillary thyroid cancer. Hum Immunol. 2013;74:447-451. 30. Carosella ED, Moreau P, Lemaoult J, Rouas-Freiss N. HLA-G: from biology to clinical benefits. Trends Immunol. 2008;29: 125-132. 31. Wastowski IJ, Sampaio-Barros PD, Amstalden EM, et al. HLA-G expression in the skin of patients with systemic sclerosis. J Rheumatol. 2009;36:1230-1234. 32. Wiendl H, Mitsdoerffer M, Hofmeister V, et al. A functional role of HLA-G expression in human gliomas: an alternative strategy of immune escape. J Immunol. 2002;168:4772-4780. Reprint requests: Aline Carvalho Batista Disciplina de Patologia Geral e Bucal Faculdade de Odontologia Universidade Federal de Goiás Praça Universitária S/N Setor Universitário CEP 74605-220 Brazil [email protected]
