Tel.: +31 10 703 3571, Fax: +31 10 703 3315
Nonsteroidal anti-inflammatory drugs do not interfere with imiquimod treatment for usual type vulvar intraepithelial neoplasia
Article first published online: 23 JUL 2010
Copyright © 2010 UICC
International Journal of Cancer
Volume 128, Issue 10, pages 2463–2469, 15 May 2011
How to Cite
Terlou, A., Kleinjan, A., Beckmann, I., Heijmans-Antonissen, C., van Seters, M., Santegoets, L. A.M., van Beurden, M., Helmerhorst, T. J.M. and Blok, L. J. (2011), Nonsteroidal anti-inflammatory drugs do not interfere with imiquimod treatment for usual type vulvar intraepithelial neoplasia. Int. J. Cancer, 128: 2463–2469. doi: 10.1002/ijc.25573
- Issue published online: 25 MAR 2011
- Article first published online: 23 JUL 2010
- Accepted manuscript online: 23 JUL 2010 12:00AM EST
- Manuscript Accepted: 16 JUL 2010
- Manuscript Received: 15 APR 2010
- vulvar intraepithelial neoplasia;
- human papilloma virus;
- nonsteroidal anti-inflammatory drugs
Imiquimod has been shown to be an effective treatment for usual type vulvar intraepithelial neoplasia (uVIN). Since local inflammation and burning are common side effects, patients often use nonsteroidal anti-inflammatory drugs (NSAIDs). Our study investigated whether NSAID-use, which has been documented to inhibit the cell-mediated immune response, interferes with the outcome of imiquimod treatment. Monocyte-derived dendritic cells (moDCs) and Langerhans cells (moLCs) were cultured in the presence of NSAIDs. The expression of relevant surface markers (CD80, CD86, CD40, HLA-DR, CCR6 and CCR7), stimulatory function, and cytokine production were evaluated. Furthermore, we analyzed in uVIN patients whether frequent NSAID-use had an effect on the clinical response and on immunocompetent cell counts before and after imiquimod treatment. Although an effect was observed on the expression of moDC and moLC maturation markers, NSAIDs did not affect the ability of moDCs and moLCs to stimulate allogeneic T-cell proliferation, or the production of cytokines in an allogeneic T-cell stimulation assay. In agreement with this, in uVIN patients treated with imiquimod, no interference of frequent NSAID-use with clinical outcome was observed. However, we did notice that high CD1a+ and CD207+ cell counts in frequent NSAID-users before treatment seemed to predict a favourable response to imiquimod treatment. Our data indicate that NSAID-use does not seem to interfere with moDC and moLC function and does not interfere with immunomodulatory properties of imiquimod in uVIN patients. Therefore, NSAIDs can safely be used to reduce imiquimod side effects in uVIN patients during treatment.
Usual type vulvar intraepithelial neoplasia (uVIN) is a premalignant skin disorder, which is caused by a persistent infection with high-risk human papilloma virus (hrHPV).1, 2 In the aetiology of uVIN, inhibition of the host immune response seems to be an important determinant. Earlier studies demonstrated that immunodeficiency and smoking (which reduces local immunity) are important risk factors for the development of HPV-related premalignant and malignant lesions.3–7
The concept that the immune response is insufficient in patients with uVIN has lead to treatment with the immune-response modifier imiquimod. Imiquimod binds to toll-like receptor (TLR) 7 and 8 at the cell surface of dendritic cells (DCs). Binding induces consecutive activation of nuclear factor-kappa B (NF-κB). This activation results in secretion of multiple cytokines, such as TNF-α, type-1 IFN and IL-12, and activation of DCs, thus inducing a strong cellular immune response.8, 9 Upon treatment of uVIN with imiquimod, 35% of patients had a complete response (100% reduction in lesion size),10 which lasted for more than 5 years in all but one case (article in preparation). In addition, another 46% of patients showed a partial reduction in lesion size of at least 25%. Only in patients where the numbers of immunocompetent cells were normalized, the clinical response was ≥75%.10 Based on these results, imiquimod is now considered first-line treatment for uVIN in the Netherlands.
Treatment with imiquimod, however, is not always tolerated well. Patients using imiquimod frequently complain of side effects such as severe local inflammation, itching or burning, flulike symptoms, weariness and headache.10, 11 We observed that patients treated with imiquimod commonly use nonsteroidal anti-inflammatory drugs (NSAIDs) or paracetamol to reduce these side effects. NSAIDs have been described to inhibit the enzymatic activity of cyclooxygenase (COX) 1 and 2, thereby impairing the conversion of arachidonic acid to prostaglandins (PGs), which results in an anti-inflammatory effect.12 Additionally, several studies demonstrated that NSAIDs also suppress NF-κB activation13–17 and in this way can inhibit DC differentiation.13, 18, 19
The effects of concomitant use of NSAIDs during imiquimod treatment are unknown. However, since the function of imiquimod is to enhance the patients' cell-mediated immune response against HPV and thus cure uVIN, the aim of our study was to investigate whether NSAIDs have an inhibitory effect on the cell-mediated immune response and thereby can interfere with imiquimods immunomodifying activities. Because it is not possible to isolate enough immune cells from VIN biopsies, we cultured DCs and Langerhans cells (LCs, a specialized dendritic cell that populates the epidermis20) from peripheral blood mononuclear cells (PBMCs) to elucidate the effect of NSAIDs on differentiation and function of antigen presenting cells. Furthermore, in patients who participated in an earlier study,10 we analyzed whether frequent use of NSAIDs had an effect on imiquimod response and on immunocompetent cell counts before and after imiquimod treatment.
Material and Methods
Isolation of CD14+ cells and culture of moDCs and moLCs
PBMCs were isolated from buffy coats of healthy donors by Ficoll-paque (Ficoll Paque Plus, GE Healthcare Biosciences, Uppsala, Sweden) centrifugation. After isolation, PBMCs were washed and CD14 microbeads (MACS Microbeads, Miltenyi Biotec, Bergisch Gladbach, Germany) were added to isolate CD14+ cells. The CD14+ fraction was counted and the monocytes were placed in 24 well plates at a density of 1 × 106 cells/ml/well. Monocytes were cultured in RPMI 1640 containing Glutamax (Gibco Invitrogen, Gibco Invitrogen, Carlsbad, CA), supplemented with 10% FCS and gentamycin, at 37°C in a humidified atmosphere containing 5% carbon dioxide. Growth factors used for the generation of monocyte-derived dendritic cells (moDCs) were 1,000 IU/ml GM-CSF (Immunotools, Friesoythe, Germany) and 200 IU/ml IL-4 (R&D systems, Abingdon, UK). TGF-β1 (10 ng/ml, PeproTech, Rocky Hill) was added to moDC cultures to generate monocyte-derived Langerhans cells (moLCs).20, 21 On day two of culture, 0.5–2.5 mM aspirin (Sigma, Zwijndrecht, the Netherlands), 100–500 μM ibuprofen (Sigma) or ethanol (Sigma) was added to the wells to measure the effect of aspirin or ibuprofen on moDC and moLC maturation. On day 6, cells were stimulated with 2 ng/ml CPG 1668 (Gibco) to mimic viral infection.
Fluorescence-activated cell sorting (FACS) analysis
On day 7, cells were harvested and the expression of membrane markers was assessed by flowcytometry using the following fluorescent labelled antibodies: CD11c APC, HLA-DR APC-Cy7 for moDC culture and HLA-DR PE-Cy7 for moLC culture (all BD Biosciences, Franklin Lakes), CD1a PE-Cy5, CD14 APC, CD80 PE (all BD Pharmingen, San Diego), CD207 PE (Immunotech, Marseille, France), CD40 APC, CD86 PE (both eBioscience, Hatfield, UK), CCR7 FITC and CCR6 FITC (both R&D systems). Cells were labelled according to the manufacturers' instructions and appropriate isotype-matched antibodies served as negative controls. After staining with the viability marker 4′,6-diamidino-2-phenylindole (DAPI, Invitrogen Molecular Probes, Carlsbad, CA), cells and data were analyzed using the LSR II (BD Biosciences) and FlowJo software (Treestar, Ashland).
Isolation and fluorescent labelling of naïve CD4+ T-cells
T-cells were purified from human buffy coat. Naïve T-cells were obtained by negative selection using naïve CD4+ T-cell isolation kit (Miltenyi) according to manufacturers' instructions. All steps were performed with MACS buffer (PBS supplemented with 5 mM EDTA and 1% BSA) at 4°C. In brief: all cells, except CD4+ naïve T-cells, were indirectly magnetically labelled with a cocktail of biotin-conjugated antibodies (CD8, CD14, CD16, CD19, CD36, CD45RO, CD123, TCRγ/δ and Glycophorin A) followed by incubation with antibiotin microbeads. Isolation of pure naïve T-cells was achieved by depletion of the magnetically labelled non-CD4+ naïve T-cells. For fluorescent cell labelling, cells were washed twice with serum-free medium and labelled in a final concentration of 5 μM carboxyfluorescein succinimidyl ester (CFSE) (Invitrogen, Molecular Probes) for 10 min in serum-free medium at 37°C. Adding excess ice-cold culture medium stopped the reaction.
Allogeneic Mixed Lymphocyte Reaction (MLR)
CFSE labelled and CFSE nonlabelled allogeneic peripheral blood naïve T-cells were co-cultured in duplicate with moDCs or moLCs with and without NSAIDs with a DC:T-cell ratio of 1:10 in culture medium in 96 well round bottom plates for 5 days at 37°C. At day 5, T-cells were harvested and stained with the antibodies CD4 APC (eBioscience) and CD3 PE (BD Pharmingen) to discriminate between T-cells and moDCs or moLCs. T-cell proliferation was measured by flowcytometry using a LSR II (BD Biosciences) and FlowJo software (Treestar).
Measurement of prostaglandin E2 and cytokine production
Levels of prostaglandin E2 (PGE2) were measured in day 7 supernatant of moDC and moLC cultures by using a competitive enzyme immunoassay (R&D systems) according to the manufacturer's instructions.
Cytokine levels were measured in MLR culture supernatants (moLCs cultured with T-cells) using ELISA kits according to the manufacturer's instructions. Il-4, IFN-γ, type 1 IFN, TNF-α (all eBioscience) and IFN-inducible protein-10 (Ip-10, BD Biosciences), were measured in day 5 supernatant of allogeneic MLR.
Study population and immunohistochemical staining of biopsies
Twenty-five women with multifocal uVIN who participated in our double-blinded, placebo-controlled randomized clinical trial (RCT)10 were treated with imiquimod twice a week for 16 weeks. Biopsy samples of these patients taken before and 4 weeks after treatment were analyzed for the presence of CD1a+ LCs, CD207+ DCs expressing Langerin, CD208+ DCs, CD123+/CD11c− plasmacytoid DCs, CD94+ natural killer (NK) cells, CD14+ monocytes, CD4+ T-helper cells, CD8+ cytotoxic T-cells, CD25+/HLA-DR+ regulatory T-cells, and CD68+ macrophages, as described previously.22
All patients kept a daily record chart to report use of their study medication as well as any concomitant medication. Drugs most commonly used in our study were over-the-counter NSAIDs. All patients who were treated with imiquimod were divided into two groups according to the following criteria for the frequency of NSAID-use: (i) Low users: never or sporadically, i.e., no more than three times during the whole study period for one or two days and (ii) frequent users: daily or regularly, i.e., more than three times during the whole study period, for 1 up to 5 days in a row.
Data analysis was performed with the use of the SPSS 15.0 software package for Windows. For the in-vitro experiments, differences in cytokine levels and in percentage of proliferated T-cells between groups were compared using the Kruskal-Wallis test.
We used the Fisher Exact test to evaluate differences in NSAID-use between patients with ≥75% clinical response, defined as a clinical reduction in lesion size of at least 75%, versus patients with <75% clinical response, defined as a clinical reduction in lesion size of less than 75%. We used the Mann Whitney U test to evaluate differences in immunocompetent cell counts between the different subgroups (frequent- or low NSAID-users with ≥75% response versus <75% response). A two-tailed p-value of p < 0.05 was chosen to represent statistical significance.
In vitro effect of aspirin and ibuprofen on moDC maturation and on moDC induced T-cell proliferation
To investigate the effect of aspirin and ibuprofen on moDC maturation, CD14+ monocytes (+ GM-CSF and IL-4) were cultured from day 2 till day 7 in the presence of 0.5–25 mM aspirin or 100–500 μM ibuprofen as described in “Material and Methods.” Since uVIN patients have a permanent HPV-infection, we added CPG on day 6 to mimic the stimulatory effect of viral DNA and detect the effect on moDC maturation and function. We performed a competitive immunoassay on day 7 culture supernatants to demonstrate that NSAIDs reduced PGE2 levels. PGE2 production was reduced by more than 20% after stimulation with ibuprofen and by more than 70% after stimulation with aspirin (data not shown).
After FACS analysis on day 7, living (DAPI-negative) moDCs (CD11c+, HLA-DR+ cells) were gated (Fig. 1a). From these viable cells, we analyzed the expression of the maturation markers CD80, CD86, CD40 and HLA-DR and of the chemokine-receptors CCR6 (highly expressed in immature DCs23) and CCR7 (highly expressed in mature DCs23). As shown in Figure 1b, a dose-dependent increase in expression of HLA-DR and CD86 was observed upon stimulation with aspirin or ibuprofen, while CD80, CD40 and CCR6 expression was somewhat decreased upon stimulation with the highest dose of aspirin or ibuprofen. No difference in expression was observed for CCR7. Similar results were observed in cell cultures without CPG stimulant (data not shown).
To investigate whether interaction of NSAID-treated moDCs with CD4+ T-cells results in T-cell activation, we co-cultured moDCs with allogeneic T-cells and subsequently analyzed T-cell proliferation. First, living CD4+CD3+ T-cells were gated and the percentage of proliferated cells was analyzed by gating the CFSE-negative T-cell population (Fig. 1c). Figure 1d shows the percentage of proliferated T-cells in co-cultures with moDCs stimulated with CPG, CPG + 2.5 mM aspirin or with CPG + 500 μM ibuprofen. MoDCs stimulated with CPG in the presence or absence of ibuprofen or aspirin did not differ significantly in their ability to activate T-cells.
Effect of aspirin and ibuprofen on moLCs
Our next step was to investigate the effect of NSAIDs on moLCs, since LCs play a major role in initiation of the cellular immune response in the skin.24 Previously, we observed that LC counts are reduced in the epidermis of VIN patients.22
To investigate the effect of NSAIDs on moLCs, we cultured PBMCs with GM-CSF, IL-4 and TGF-β to generate a high percentage of Langerhans-like moDCs. As is shown in Figure 2a, CD1a—a cell-specific marker for epidermal LCs24—is expressed on the surface of cells cultured with TGF-β. However, upon stimulation with aspirin or ibuprofen, more cells became CD1a negative suggesting an impaired differentiation. The strongest effect was observed after stimulation with ibuprofen. The expression of CD207—Langerin, which is a LC specific C-type lectin24—decreased upon aspirin stimulation but increased upon ibuprofen stimulation. When we measured the expression of the maturation markers CD80, CD86 and HLA-DR, an increase in CD86 expression was found upon stimulation with aspirin or ibuprofen (Fig. 2a), as was observed in conventional moDCs. However, HLA-DR expression increased upon stimulation with ibuprofen, but decreased upon stimulation with aspirin. CCR6 increased upon ibuprofen stimulation and no differences were found for the other markers (Fig. 2a).
To assess the effect of NSAID-stimulated moLCs on T-cell proliferation, we performed again an allogeneic MLR. However, moLCs cultured with or without ibuprofen or aspirin did not differ significantly in their ability to activate T-cells (Fig. 2b).
The levels of different inflammatory cytokines were measured in supernatant of MLR cultures to assess the effect of NSAIDs on cytokine production. Although Ip-10 and Il-4 were significantly higher in MLR supernatants than in T-cells cultured without moLCs, no effect of aspirin or ibuprofen was observed on cytokine production (Fig. 2c).
In vivo effect of NSAID-use on immunocompetent cell counts before and after imiquimod treatment, and response rate after treatment
Results from our in-vitro experiments demonstrated no effect of NSAIDs on phenotype and function of moDCs and moLCs. Because we were interested in a possible interference of NSAIDs with imiquimod treatment, we next investigated the effect of NSAID-use on imiquimod response in VIN patients who were treated with imiquimod in a previous double-blinded, placebo-controlled RCT.10 According to our definition, 12 of 25 patients were frequent NSAID-users; five of the 12 had a clinical response to imiquimod of ≥75% and seven patients had a response of <75%. Thirteen patients were low NSAID-users; eight of them had a clinical response of ≥75% response and five patients had a response of <75%. No significant difference was measured in NSAID-use between patients with ≥75% response and patients with <75% response. This indicates that NSAID-use does not influence the success of imiquimod treatment. Interestingly, when we compared immunocompetent cell counts between different subgroups, we did observe that frequent NSAID-users with ≥75% response had significantly higher CD1a+ and CD207+ cell counts in the epidermis before imiquimod treatment than frequent NSAID-users with <75% response (p = 0.007 and p = 0.004 respectively, Figs. 3a and 3b) or than any of the low NSAID-users. Cell counts for the different immunocompetent cells are shown in Supporting Information Table 1.
In conclusion, elevated CD1a+ and CD207+ cell counts in frequent NSAID-users seem to predict a good clinical response to imiquimod treatment, but NSAID-use, as such, does not negatively affect the success of imiquimod treatment.
In our study, we investigated whether NSAIDs have an inhibitory effect on the cell-mediated immune response and thereby can interfere with imiquimods immunomodifying activities. For this purpose, moDCs and moLCs were cultured in vitro in the presence of aspirin and ibuprofen and expression of relevant surface markers, stimulatory function and production of pro-inflammatory cytokines were measured. Although an effect was observed on the expression of moDC and moLC maturation markers (HLA-DR, CD86, CD80, CD40 and CCR6), NSAIDs did not affect the ability of moDCs and moLCs to stimulate T-cells, neither could we observe an effect on the production of cytokines. Additionally, we investigated whether frequent use of NSAIDs had an effect on imiquimod response in patients with uVIN. We found that frequent NSAID-users with high CD1a+ and CD207+ cell counts before imiquimod treatment had a favourable response to imiquimod treatment. In other words, in frequent NSAID-users high CD1a+ and CD207+ cell counts seem to predict the imiquimod response. However, in accordance with our in vitro findings, we did not observe a reduced imiquimod response in VIN-lesions of frequent NSAID-users. In summary, our results indicate that NSAIDs do not inhibit moDC and moLC stimulatory function and do not seem to interfere with imiquimods immunomodulatory properties in uVIN patients.
NSAIDs have been used for more than 100 years to reduce pain and inflammation, in particular in patients with various rheumatologic and inflammatory disease states for adequate control of their condition. The anti-inflammatory and analgesic efficacy of NSAIDs is mainly due to the inhibition of COX-1 and COX-2, which convert arachidonic acid to PGs.12 PGs play an important role in the function of immune cells by modulating production of cytokines, chemokines and their receptors.25 Additionally, PGs such as PGE2 are mandatory for DC migration26, 27 and enhance T-cell proliferation and differentiation.28
Another way in which NSAIDs affect the immune response is by inhibiting NF-κB activation.13 NF-κB has been regarded as a key element in the response of cells to inflammatory stimuli and targets pro-inflammatory cytokines, chemokines and cell adhesion molecules.15 Inhibition of NF-κB also affects maturation of DCs.29 Several studies have investigated the effect of NSAIDs on DCs.18, 19, 30–34 Hackstein et al.18 reported an inhibition of maturation and function of murine DCs after stimulation with aspirin, while Matasic et al.19 also demonstrated an impairment of immunostimulatory function in human DCs after aspirin stimulation. This effect was due to inhibition of NF-κB, since no effect was observed upon stimulation with selective COX-1 or COX-2 inhibitors.19 In contrast, other studies demonstrated that NSAIDs can stimulate DC function by inhibition of PGE2 production. PGE2 enhances Il-10, which down-regulates Il12-p70 production and thereby down-regulates DC stimulatory capacity.33, 34 These sometimes contradictory reports are difficult to compare, since different cell types, protocols or drugs were used. Moreover, various pathways involved in DC maturation can be affected upon stimulation with NSAIDs.
In our study, no effect on moDC function was measured after stimulation with aspirin or ibuprofen, despite the fact that the concentrations of aspirin and ibuprofen used in our experiments were within the range to inhibit COX-1, COX-2 and NF-κB activation.16, 17 In addition, we also failed to observe an effect of NSAID-use during imiquimod treatment of VIN. A possible explanation for this is that a significant percentage of NSAIDs bind to plasma albumin, and in vivo NSAIDs are partly converted into more or less biologically active metabolites,15, 35 which could result in plasma concentrations of NSAIDs in patients too low to affect DC function. It is also possible that the effect of NSAIDs in these patients is overruled by the strong stimulatory effect of imiquimod on NF-κB. After binding on TLR 7 or 8 on the surface of the DC, imiquimod induces activation of NF-κB, which results in a profound cellular immune response.8, 9
In conclusion, our data indicate that NSAIDs do not interfere with imiquimod immunomodifying activities and therefore can safely be used to reduce the symptoms of imiquimod side effects in uVIN patients during treatment. Furthermore, elevated CD1a+ and CD207+ cell counts in frequent NSAID-users seem to predict a good clinical response to imiquimod treatment.
Additional Supporting Information may be found in the online version of this article.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.