Characterizing T-cell response in low-grade and high-grade vulval intraepithelial neoplasia, study of CD3, CD4 and CD8 expressions

Gynecologic Oncology 94 (2004) 48 – 53 www.elsevier.com/locate/ygyno

Characterizing T-cell response in low-grade and high-grade vulval intraepithelial neoplasia, study of CD3, CD4 and CD8 expressions Nahid Gul, a,* Raji Ganesan, b and David M. Luesley a a

Department of Gynaecological Oncology, Birmingham Women’s Hospital, Edgbaston, Birmingham, UK b Department of Histopathology, City Hospital, Birmingham, UK Received 19 August 2003 Available online 18 May 2004

Abstract Objective. The objective of our study was to compare immunocyte infiltrates in vulval epithelium from low-grade and high-grade vulval intraepithelial neoplasia (VIN) lesions to determine if difference in T-cell presence reflected the grade of VIN. Material and methods. Thirty-six vulval specimens were obtained from 24 patients who had previously undergone vulval biopsies for VIN, 14 high-grade diseases (VIN 3 with or without HPV) and 14 low-grade diseases (VIN 1 and VIN 2 with or without HPV). Eight samples of normal vulval tissue were selected from the excision margins of resected vulval biopsies. The lymphocyte surface markers included CD3 (Pan T-cell marker), CD4 (T helper cells), and CD8 (T cytotoxic cells). Each tissue section was visualized under high power magnification and cells were counted in 10 random areas at the dermo-epidermal junction. Results. A significantly higher number of total mean T lymphocytes were detected in VIN specimens compared to normal vulval tissue ( P = 0.002). In low-grade VIN, there were significantly more CD8 cells than CD4 when compared to high-grade VIN. This difference in CD4/CD8 ratio was significant ( P = 0.001). Conclusions. This study suggests that increased CD8 response in VIN is a feature of low-grade disease and we speculate that this may be a protective mechanism. In high-grade disease, both CD4 cells and CD8 cells are equally present with preservation of normal CD4/CD8 ratio. D 2004 Elsevier Inc. All rights reserved. Keywords: CD subsets; T helper; T cytotoxic; VIN

Introduction The natural history of vulval intraepithelial neoplasia (VIN) is not clear. Morphologically and histologically identical lesions behave differently. Progression of VIN 3 to invasive carcinoma is variable and is reported in 1– 10% of cases. Epithelium adjacent to squamous cell carcinoma (SCC) of the vulva shows morphological changes of VIN or HPV with concomitant VIN 3 reported in 22– 67% cases. Lichen sclerosis in nearby epithelium is seen in 10– 62%. This suggests the presence of some common local factors predisposing to the premalignant and malignant lesions. Increased risk of malignancies has been reported in patients with immune deficiency. The best examples of these are

* Corresponding author. Cornerways, 13 The Spinney, Gayton, Wirral, CH60 3SU, UK. Fax: +44-151-6041552. E-mail address: [email protected] (N. Gul). 0090-8258/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2004.02.037

transplant patients on immunosuppressive drugs and patients with HIV infection. No data exist to demonstrate T-cells and subsets of T-cells in vulval preneoplastic disease or vulval cancer. The T-cell response is studied by total Tcell count (CD3) and the T-cell subsets (CD4 helper and CD8 cytotoxic cells). Vulval epidermal T-cells are mainly CD8; however in the dermis, CD4 and CD8 cells are equally distributed [1]. This proportion of T-cell subset is different than peripheral blood where the ratio of CD4/CD8 is normally 2:1. Changes in the local immune responses in the vulva may have implications for the development of different grades of VIN. We hypothesize a role for altered local immune response in determining the progression of preneoplasia of the vulva. In the present study, we investigated T-cell subsets as determined by CD antigen presentation, in lowgrade and high-grade vulval intraepithelial neoplasia. We studied the antigen presentation by immunohistochemical staining of archival vulval tissue.

N. Gul et al. / Gynecologic Oncology 94 (2004) 48–53

Material and methods Patients Thirty-six histology specimens from 24 women who had previously undergone vulval biopsies for VIN were taken for the present study from the laboratory archives. The selected cases were consecutive samples identified through the histopathology database with histological features of VIN. The database was searched for cases of VIN from January 1999 to June 1999. During this period, 14 cases of low-grade VIN (VIN 1 and VIN 2) were identified. Fourteen consecutive cases of VIN 3 during the same period were selected for the high-grade VIN group. Eight specimens of vulva with no VIN or HPV changes were used as normal tissues. These were selected from the excision margins of vulval specimens showing no morphological abnormality. No cases of differentiated VIN were included and cases of invasive carcinoma were excluded from this study. All histological diagnoses were confirmed by a single pathologist (RG) to eliminate inter-observer variation. Demographic characteristics including age, symptoms, presence of immunodeficiency, smoking, history of warts, CIN, and HPV were abstracted from individual case notes. Study samples were taken from paraffin sections of the identified biopsy specimens. Immunohistochemical studies for CD3, CD4, and CD8 were undertaken using primary antiCD3, anti-CD4, and anti-CD8 antibodies, respectively. Serial sections were stained with CD3, CD4, and CD8 antibodies allowing comparison in the same microscopic area.

49

washes in PBS was followed by incubation for 5 – 10 min in DAB solution (Vector Laboratories, Peterborough) to visualize the antigen. A thorough wash in running tap water was followed by light staining of the nuclei and ‘‘bluing’’ in Scott’s water substitute. Sections were then taken back to xylene and mounted. Cell count The cells positive for CD3, CD4, and CD8 were determined by qualitative and quantitative staining in histological sections at the dermo-epidermal junction of the vulva. The intensity and distribution of positive cells were recorded and scored from 1 to 4: 1—scanty; 2—few superficial; 3— positive patchy; 4—positive continuous. For quantitative assessment, all positive cells per mm2 of vertical epithelial sections at the dermo-epidermal junction in 10 high-power fields (400) for each antibody were counted using the calibrated grid in the eyepiece. Data were expressed as a mean of cells/10 high-power fields. The slides were independently examined by two authors (NG and RG). Statistical analysis The data were analyzed using SPSS version 10 (Statistical Package for Social Sciences). The mean F SD was estimated for each antibody. Differences between the groups were evaluated using one-way ANOVA test. A P value of <0.05 was regarded as significant.

Immunohistochemical staining

Results

Three-micrometer sections from archived tissue blocks were cut onto APES-coated slides (Merck, Leicestershire) and allowed to dry at room temperature for approximately 30 min. Sections were then incubated at 60jC for 2 h. Sections of normal human tonsils were used as positive control and also run in parallel with the test sections to validate the technique. Slides were washed in distilled water for 10 min and endogenous peroxidases blocked by using a 5% solution of hydrogen peroxide (Merck). Antigen unmasking was achieved using microwaving techniques. Sections were washed in running tap water and then placed in a 10% solution of normal swine serum (DAKO, Cambridgeshire) to reduce nonspecific background staining. The primary antibody of 1 in 200 dilution for CD3 (DAKO), 1 in 10 for CD4 (Ventana), and 1 in 40 for CD8 (Novocastra) in phosphatebuffered saline (PBS) at pH 7.5 was then overlaid onto the sections for 1 h at room temperature in humidity chambers. The primary antibody was washed off using three washes with PBS. Incubation with the secondary link reagent was then carried out using a LSAB II large volume kit (DAKO). Following 10 min incubation at room temperature, the sections were washed in PBS and overlaid with the tertiary label reagent from the same source for 10 min. Three

Demographic characteristics The 36 specimens included low-grade VIN 14 (38%) and high-grade VIN 14 (38%) and 8 (22%) normal epithelium. The mean age of the patients was 45 years with a range of 17 – 88. The commonest symptom was pruritus noted in 67%. Fifty-four percent were smokers, 12% had a history of immunodeficiency, and 33% had a history of previous CIN. Thirty-seven percent had a history of previous genital warts. Table 1 summarizes the possible confounders of the study. The influence of putative cofactors on CD4/CD8 ratio in VIN was tested in a multiple regression model and presence of HPV was the only significant factor ( P = 0.030), while age, smoking, symptoms, immunodeficiency, warts, and history of CIN were not significant. Table 1 Possible confounders to the study Factors

Percentage (number)

Immunodeficiency Smokers History of genital warts History of CIN

12% 54% 37% 33%

(3) (13) (9) (8)

50

N. Gul et al. / Gynecologic Oncology 94 (2004) 48–53

Histopathological findings

Table 2 Density of different T-cell subsets and CD4/CD8 ratios in all cases

Twenty-eight specimens had histologically confirmed VIN of different grades. Six specimens of VIN 1 mainly showed features of HPV and probable VIN 1. Nevertheless, as inter-observer variation is well known especially in diagnosing VIN 1 and HPV is known to play a significant role in VIN, we included these in the low-grade disease group. Features of human papilloma virus were histologically diagnosed by the presence of koilocytes and multinucleation. Fourteen (38%) cases of low-grade VIN included VIN 1 and VIN 2 with or without HPV. Fourteen (38%) cases of high-grade VIN included VIN 3 with or without HPV. Histologically, HPV was diagnosed in 58% of specimens. The primary aim of this study was not related to HPV; hence, HPV typing was not undertaken.

Serial no.

Case no.

Diagnosis

CD3

CD4 per CD3 (%)

CD8 per CD3 (%)

CD4/CD8 ratios

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

27 28 30 33 34 35 36 20 21 1 4 5 12 14 9 11 13 26 23 24 3 10 31 32 2 8 18 19 6 22 25 7 29 15 16 17

normal normal normal normal normal normal normal normal HPV/VIN 1 HPV/VIN 1 HPV/VIN 1 HPV/VIN 1 HPV/VIN 1 HPV/VIN 1 VIN 1 VIN 1 HPV VIN 1 HPV VIN 1 HPV VIN 2 VIN 2 VIN 2 HPV VIN 2 HPV VIN 3 VIN 3 VIN 3 VIN 3 VIN 3 VIN 3 VIN 3 VIN 3 VIN 3 HPV VIN 3 HPV VIN 3 HPV VIN 3 HPV VIN 3 HPV VIN 3 HPV

8 32 23 7 8 10 7 7 22 191 132 21 22 15 36 10 25 13 32 54 86 23 34 42 30 41 23 20 76 41 91 53 40 22 27 66

50 22 17 43 50 50 50 57 45 7 3 38 23 33 39 30 36 46 44 48 43 48 50 48 50 44 52 50 44 49 52 49 25 45 48 47

50 78 83 57 50 50 50 43 55 62 91 62 82 60 61 70 60 46 53 52 45 52 53 50 50 44 50 45 51 49 48 51 75 55 52 45

1.0 0.28 0.20 0.75 1.0 1.0 1.0 1.0 0.81 0.11 0.03 0.61 0.27 0.55 0.63 0.42 0.60 1.0 0.83 0.92 0.94 0.91 0.94 0.96 1.0 1.0 1.0 1.0 0.87 1.0 1.08 0.96 0.33 0.81 0.92 1.0

Quantitative T-cell count The density of the different T-cell subsets is presented in Table 2. The results of different groups are summarized in Table 3. The mean density of CD3 cells was 12 F 3 cells/ mm2 in normal vulval tissue, 48 F 14 cells/mm2 in lowgrade VIN, and 43 F 6 cells/mm2 in high-grade VIN. In the ‘‘normal’’ group, the total T-cell count was lower than in VIN; however, there was no difference in the subset of Tcells. The difference in the mean T-subsets between the three groups is shown in Table 4. In low-grade VIN, the CD8 count was significantly higher than CD4 positive cells with a mean CD4/CD8 ratio of 0.54. The mean CD4/CD8 ratio in high-grade VIN was 0.91, with no significant difference in CD4 and CD8 positive cells. The mean CD4/CD8 ratio was significantly decreased in low-grade as compared to high-grade VIN ( P = 0.000) (Table 5).

Mean cell number per ten 400 fields.

Qualitative lymphocyte count Total T-lymphocytes were detected by CD3 antibody staining and were distributed in continuous or patchy distribution at the dermo-epidermal junction. In normal vulval tissue, the T-cell infiltrate was scanty and subepidermal (Figs. 1A, B), while in low-grade VIN, 43% had patchy distribution of T-cells at dermo-epidermal junction. In the majority of cases of high-grade vulval intraepithelial neoplasia, the T-cell infiltrate was continuous mainly in the dermis close to the epidermal – dermal junction. The Tsubset was mostly CD8 in low-grade VIN (Figs. 2A, B) and equal distribution of CD4 and CD8 in high-grade disease (Figs. 3A, B).

Discussion There are no data available to demonstrate the natural history of VIN 1 and VIN 2. We placed VIN 1 and VIN 2 into low-grade VIN as the risk of malignancy almost

certainly is low [2], while progression of VIN 3 to invasive squamous cell carcinoma has been well demonstrated [3]. The risk of progression is variable particularly in older women ranging from 3% to 19% [4]. Moreover, VIN 3 may be present in epithelium adjacent to invasive squamous cell carcinoma in 20– 30% of cases [5]. Clearly from the available evidence, it would appear that low-grade and highgrade VINs behave differently. Our results show a different T-cell subset response in the two grades of VIN. At least two distinct types of vulvar squamous cell carcinoma are well recognized, one in younger patients with HPV infection and one in older patients without HPV infection [4]. Our study included patients from younger and older age groups; however, there was no significant difference in the subsets of T-cells and CD4/CD8 ratio with age. It has been shown that the inability of the host to reject a tumor may be due to insufficient generation of tumor specific CD4 T-cells [6,7]. Various reports in the literature have suggested decreased CD4/CD8 ratios in tumor-infil-

N. Gul et al. / Gynecologic Oncology 94 (2004) 48–53

51

Table 3 Number of T-Cell subsets in epidermal – dermal junction of vulva*

Table 5 Comparison of CD4/CD8 ratio

Groups

No. of patients

CD3

CD4

CD8

Groups

No.

CD4/CD8 ratio

P value

Normal Low-grade VIN High-grade VIN

8 14

12.75 F 3.35 48.36 F 14.13

6.13 F 1.56 10.29 F 1.65

6.75 F 1.90 33.07 F 10.42

Normal Low-grade VIN High-grade VIN

8 14 14

0.96 0.54 0.91

0.000

14

43.29 F 5.70

20.12 F 2.90

22.07 F 2.82

* Mean cell number per ten 400 field F standard error.

trating lymphocytes as an indicator of disease progression [8,9]. While cytotoxic CD8 lymphocytes are shown to have the ability for lysing cancer cells that express the tumorspecific antigen, this process may vary in different tissues and tumor types. Unlike cervical tissue, normal vulval tissue is not easily available for research. We used normal epithelial tissue from the resection margins of the vulval specimens taken from patients who had undergone vulval biopsies for vulval carcinoma or other benign vulval diseases. While normality of these tissue can be questioned, Kagie et al. [10] used similar tissue for controls in their study and found HPV on PCR studies in only 4% of such cases. Others have used the skin from posterior vaginal repairs for comparison as normal controls. The specimen obtained after such a procedure is skin at the fourchette that is vaginal epithelium and sometimes the inner aspect of labia minora that is noncornified and hence may be different to the majority of cornified vulval tissue. The main purpose of our study was to study the T-cell response in different grades of VIN; hence, we used histologically normal tissues with no morphological features of VIN as the normal group. The T-cell distribution and CD4/CD8 ratio in this group was what would be expected in a normal skin epidermal/dermal tissue [1]. Etiology of VIN is multifactorial with an assumed role for HPV and presumptive role of other cofactors. Comparing the CD4/CD8 ratio and the cofactors for VIN, presence of HPV as defined microscopically was the only influencing factor. Cigarette smoking is thought to be one of the cofactors for VIN [11]. In our study, 54% were smokers; however, the CD4/CD8 ratio did not change with smoking status ( P = 0.372). The total number of T-cells was not significantly different in low-grade and high-grade VINs and the total number of CD8 cells appeared similar in both groups. However, the change in CD4/CD8 ratio was due to a relative increase in CD8 compared to CD4 cells in low-grade VIN. The other

reason for a decreased CD4/CD8 ratio could be a decreased CD4 count. The total number of CD4 cells was not decreased when compared to normal vulval tissue. Therefore, the difference in CD4/CD8 ratio between low- and highgrade VINs is most likely to be the result of increased CD8 cells in low-grade VIN. The mean CD4/CD8 ratio was 0.91 in high-grade VIN with equal numbers of T helper and T cytotoxic cells. The absence of increased CD8 cells as compared to CD4 cells in high-grade VIN is similar to studies showing that tumor-specific CD8 cytotoxic cells

Table 4 Comparison of the mean T-cell subsets Antigen

Normal

LG/VIN

HG/VIN

P value

CD3 CD4 CD8

12 6 6

48 10 33

43 20 23

0.083 0.001 0.080

LG/VIN = low-grade vulval intraepithelial neoplasia, HG/VIN = high-grade vulval intraepithelial neoplasia.

Fig. 1. Immunohistochemical detection of CD4 immunocytes (A) and CD8 immunocytes (B) in normal vulval epithelium.

52

N. Gul et al. / Gynecologic Oncology 94 (2004) 48–53

primary aim of our study. Increased cytotoxic CD8 response was seen in vulval tissue with HPV features without VIN. However, the difference in T-cell subset in low- and highgrade VIN appear to be independent of koilocytosis ( P = 0.215). In this study, low- and high-grade VINs were associated with significantly different CD4/CD8 ratios. We feel these findings have clinical significance and propose that an increase in CD8 cells is the first local immune response in VIN. This protective cytotoxic CD8 activity is lost as VIN progresses. A similar increase in CD8 response has been shown in cancer patients after treatment with tumor necrotic factor (TNF) [14]. Protective activity of cytotoxic CD8 response has also been demonstrated against fungal infections in experimental studies [15]. The other explanation for increased CD8 cells in low-grade VIN may be due to the presence of memory T-cells (CD45RO), or that the T-cells which are present in high-grade VIN are functionally inactive. The findings of our study suggest that in lowgrade VIN, there appears to be a relatively normal local

Fig. 2. Immunohistochemical detection of CD4 immunocytes (A) and CD8 immunocytes (B) in low-grade VIN.

often disappear before a tumor is completely eliminated [12]. This is thought to be due to tumor cells inducing apoptosis of CTL through CD95 –CD95 ligand interaction. Decreased CD4/CD8 ratios have been shown in cervical cancer due to the significantly lower CD4 cell count [9]. Unlike this observation, the decrease in CD4/CD8 ratio in our low-grade VIN group was due to increased CD8 cytotoxic cells rather than decreased CD4 cells. The decrease in CD4 cells may also be the a feature of neoplastic process. The other explanation for the different T-cell response may be due to the difference in the etiology of vulval and cervical neoplasia. HPV is associated with the majority of squamous cell carcinomas of cervix unlike vulval carcinoma. Squamous epithelium is mostly keratinized in the vulva and nonkeratinized in the cervix. HPV is associated with invasive cancer in epithelium not showing keratinization [13]. We found histological features of HPV in 54% of VIN 3, similar to other reports in the literature [10]. The role of HPV can only be addressed in studies using sensitive methods of HPV detection and typing, which was not the

Fig. 3. Immunohistochemical detection of CD4 immunocytes (A) and CD8 immunocytes (B) in high-grade VIN.

N. Gul et al. / Gynecologic Oncology 94 (2004) 48–53

immune response; hence, conservative approaches of management may be justifiable. The best measure of T-cell response should include studies undertaking stimulation of T-cells in vitro. However, such studies require fresh vulval specimens. We have undertaken preliminary immunohistochemical studies that suggest differences in T-subsets in different grades of VIN.

Conclusion Our study demonstrates difference in the T-lymphocytes in different grades of vulval intraepithelial neoplasia. The total mean T-Cell count appears higher in VIN as compared to normal vulval tissue. However, the CD4/CD8 ratio was decreased in low-grade disease compared to high-grade disease and ‘‘normal’’ tissue. This difference in CD4/CD8 ratios in two grades of VIN was mostly due to increased cytotoxic T-cells (CD8) cells in low-grade VIN as compared to high grade. This immune response may be a protective feature preventing disease progression. However, this can only be explained if the various grades of VIN are continuum of the same disease. The relative lack of increased CD8 cells than CD4 cells in VIN 3 compared to low grade may be due to lack of costimulatory factors or presence of inactive T-cells. Further studies are required to fully evaluate these findings.

[4] [5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

References [14] [1] Abbas AK, Lichtman AH, Pober JS. Cells and tissues of the immune system. Cellular and molecular immunology. Fourth edition 2000; 17 – 38 (Chap. 2). [2] Bancher-Todesca D, Obermair A, Bilgi S, Kohlberger P, Kainz C, Bretenecker G, et al. Angiogenesis in vulvar intraepithelial neoplasia. Gynecol Oncol 1997;64:496 – 500. [3] Herod JJO, Shafi MI, Rollason TP, Jordan JA, Luesley DM. Vulvar

[15]

53

intraepithelial neoplasia: long term follow up of treated and untreated women. Br J Obstet Gynaecol 1996;103:446. Crum CP. Carcinoma of the vulva: epidemiology and pathogenesis. Obstet Gynecol 1992;79:448 – 58. Leibowitch M, Neill S, Pelisse M, Moyal-Barracco M. The epithelial changes associated with squamous cell carcinoma of the vulva: a review of the clinical, histological and viral findings in 78 women. Br J Obstet Gynaecol 1990;97:1135 – 9. Greenberg PD. Adoptive T cells therapy of tumors: mechanisms operative in the recognition and elimination of tumor cells. Adv Immunol 1991;49:281 – 355. Basker S, Glimcher L, Nabavi N, Jones RT, Ostrand-Rosenberg S. Major histocompatibility complex 11+B7 1+ tumor cells are potent vaccines for stimulating tumor rejection on tumor-bearing mice. J Exp Med 1995;181:619 – 29. Igarashi T, Murakami S, Takahashi H, Matsuzaki O, Shimazaki J. Changes on distribution of CD4+/CD45RA and CD8+/CD11 cells in tumor-infiltrating lymphocytes of renal cell carcinoma associated with tumor progression. Eur Urol 1992;22:323 – 8. Sheu B-C, Hsu S-M, Ho H-N, Lin R-H, Torng P-L, Huang S-C. Reversed CD4/CD8 ratios of tumor-infiltrating lymphocytes are correlated with the progression of human cervical carcinoma. Cancer 1999;86(8):1537 – 43. Kagie MJ, Kenter GC, Zomerdijk-Nooijen Y, Hermans J, Schuuring E, Timmers P, et al. Human papillomavirus infection in squamous cell carcinoma of the vulva, in various synchronous epithelial changes and in normal vulva skin. Gynecol Oncol 1997;67:178 – 83. Wilkinson EJ, Cook JC, Friedrich EG, Massey JK. Vulvar intraepithelial neoplasia: association with cigarette smoking. Colposcopy Gynecol Laser Surg 1988;4:153 – 9. Contreras DN, Krammer PH, Potkul RK, Bu P, Rossi JL, Kaufmann AM, et al. Cervical cancer cells induce apoptosis of cytotoxic T lymphocytes. J Immunother 2000;23(1):67 – 74. Nuovo GJ, Delvenne P, MacConnell P, Chalas E, Neto C, Mann WJ. Correlation of histology and detection of human papillomavirus DNA in vulvar cancers. Gynecol Oncol 1991;43:275 – 80. Braczkowski R, Zubelewicz B, Romanowski W, Lissoni P, Brivio F. Possible biological indicators of the response to recombinant tumor necrosis factor alpha in patients with advanced neoplasms. J Exp Clin Cancer Res 1998;17(3):349 – 54. Cano LE, Singer-Vermes LM, Costa TA, Mengel JO, Xidieh CF, Arruda C, et al. Depletion of CD8(+) T cells in vivo impairs host defense of mice resistant and susceptible to pulmonary paracoccidioidomycosis. Infect Immun 2000;68(1):352 – 9.