Nimesulide, a COX-2 inhibitor, does not reduce lesion size or number in a nude mouse model of endometriosis

M.L. Hull1,4, A. Prentice2, D.Y. Wang1, R.P. Butt3, S.C. Phillips3, S.K. Smith1 and D.S. Charnock-Jones1,2

1 Reproductive Molecular Research Group, Department of Pathology Tennis Court Road, Cambridge CB2 1QP, 2 University Department of Obstetrics and Gynaecology, The Rosie Hospital, Robinson Way, Cambridge CB2 2SW and 3 Discovery Biology, Pfizer Global Research and Development, Sandwich, Kent, UK

4 To whom correspondence should be addressed. Email: mlh30{at}cam.ac.uk


    Abstract
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: Women with endometriosis have elevated levels of cyclooxygenase-2 (COX-2) in peritoneal macrophages and endometriotic tissue. Inhibition of COX-2 has been shown to reduce inflammation, angiogenesis and cellular proliferation. It may also downregulate aromatase activity in ectopic endometrial lesions. Ectopic endometrial establishment and growth are therefore likely to be suppressed in the presence of COX-2 inhibitors. We hypothesized that COX-2 inhibition would reduce the size and number of ectopic human endometrial lesions in a nude mouse model of endometriosis. METHODS: The selective COX-2 inhibitor, nimesulide, was administered to estrogen-supplemented nude mice implanted with human endometrial tissue. Ten days after implantation, the number and size of ectopic endometrial lesions were evaluated and compared with lesions from a control group. Immunohistochemical assessment of vascular development and macrophage and myofibroblast infiltration in control and treated lesions was performed. RESULTS: There was no difference in the number or size of ectopic endometrial lesions in control and nimesulide-treated nude mice. Nimesulide did not induce a visually identifiable difference in blood vessel development or macrophage or myofibroblast infiltration in nude mouse explants. CONCLUSION: The hypothesized biological properties of COX-2 inhibition did not influence lesion number or size in the nude mouse model of endometriosis.

Key words: COX-2 inhibitor/endometriosis/mouse model/nimesulide/prostaglandin


    Introduction
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Endometriosis affects 10% of women of reproductive age, inducing symptoms of chronic pelvic pain, dysmenorrhoea, dyspareunia and subfertility (Strathy et al., 1982Go). Non-steroidal anti-inflammatory (NSAID) medication is a longstanding treatment for the symptomatic relief of endometriosis-associated pain (Mahutte and Arici, 2003Go). The effect of NSAIDs on endometriotic lesion size and number has not been evaluated in clinical trials due to the risks and morbidity associated with repetitive laparoscopy. We theorized that inhibition of the prostaglandin pathway was likely to reduce the survival and growth of endometrium at an ectopic site.

Cyclooxygenases (COX) regulate the rate-limiting conversion of arachidonic acid to prostaglandin H2 (PGH2) in the prostaglandin synthesis pathway (Smith et al., 1996Go). Unlike COX-1 that is constitutively expressed, COX-2 transcription is upregulated at sites of inflammation (Vane et al., 1994Go). COX-2 has been implicated in inflammatory, angiogenic, proliferative and estrogenic cellular processes (Dannenberg et al., 2001Go) and is likely to be influential in several of the processes that lead to ectopic endometrial development. Inflammatory mediators such as tumour necrosis factor-{alpha} (TNF-{alpha}) and interleukin-1{beta} (IL-1{beta}) upregulated COX-2 mRNA and protein expression in cultured peritoneal macrophages (Wu et al., 2002Go). These factors induced the phosphorylation of inhibitor-{kappa}B and promoted the nuclear translocation of nuclear factor-{kappa}B (NF-{kappa}B) in endometriotic stromal cells (Sakamoto et al., 2003Go). The NF-{kappa}B binding region of the COX-2 promoter gene was found to be essential for COX-2 promoter activity in endometrial stromal cells (Tamura et al., 2002aGo,bGo). Other signalling pathways may also participate in IL-1{beta}-mediated COX-2 regulation in endometrial stromal cells as inhibitors of extracellularly-related kinases 1 and 2, protein kinase A (PKA) as well as NF-{kappa}B were shown to inhibit IL-1{beta}-induced COX-2 upregulation (Tamura et al., 2002aGo). Both IL-1{beta} (Tamura et al., 2002aGo) and PGE2 (the main prostaglandin product of the COX pathway) (Tamura et al., 2002bGo) acted post-transcriptionally to stabilize COX-2 transcripts promoting sustained COX-2 enzyme activity in an inflammatory environment.

Endometriosis is an inflammatory condition (Haney and Weinberg, 1988Go). PGE2 (Karck et al., 1996Go; Raiter-Tenenbaum et al., 1998Go), TNF-{alpha} and IL-1{beta} (Mori et al., 1991Go; Keenan et al., 1995Go) are all elevated in peritoneal macrophage-conditioned media or in the peritoneal fluid from women with endometriosis when compared with disease-free women. The peritoneal macrophages of women with endometriosis have higher levels of COX-2 mRNA and protein than those from women without the disease (Wu et al., 2002Go). COX-2 mRNA and protein are also present in higher amounts in ectopic endometrial lesions when compared with eutopic endometrium from women with endometriosis. Eutopic tissue from disease-free women had the lowest levels of COX-2 mRNA and protein (Ota et al., 2001Go; Chishima et al., 2002Go). We hypothesized that a reduction in prostaglandin levels secondary to COX-2 inhibition was likely to disrupt the inflammatory processes evident in women with endometriosis.

Endometriotic blood vessel formation may be disrupted by the restriction of COX-2 activity. Products of the COX pathway, PGE1 and PGE2 (Diaz-Flores et al., 1994Go; Form and Auerbach, 1983Go; Milne and Jabbour, 2003Go), and thromboxane A2 (Daniel et al., 1999Go; Nie et al., 2000Go) promote blood vessel formation in vitro and in vivo. Additionally, COX-2 inhibitors reduce the vascularity of in vivo angiogenesis (Masferrer et al., 2000Go) and tumorigenesis models (Tsujii et al., 1998Go; Dormond et al., 2001Go). NS-398, a specific COX-2 inhibitor, was shown to inhibit endothelial cell spreading and migration in vitro by suppressing {alpha}V{beta}3 activation of the cdc42 and Rac GTPases as well as reduce fibroblast growth factor-mediated angiogenesis in vivo (Dormond et al., 2001Go).

There is evidence that COX-2 inhibitors also inhibit vascular endothelial growth factor-A (VEGF-A)-mediated angiogenesis PGE2 induced VEGF expression in synovial fibroblasts in vitro (Ben-Av et al., 1995Go). COX-2-overexpressing colonic cancer cells secreted high levels of VEGF-A (Tsuji et al., 1998Go) and a significant association between COX-2, VEGF and microvascular density was demonstrated in human head and neck cancer (Gallo et al., 2002Go). COX-2 (–/–) mice have abnormal implantation and decidualization associated with defective VEGF-A signalling and abnormal angiogenesis in the endometrium (Matsumoto et al., 2002Go). The importance of VEGF-A-mediated angiogenesis in ectopic endometrial lesion formation has been demonstrated in a nude mouse model of endometriosis (Hull et al., 2003Go). We postulated that inhibition of COX-2 would reduce VEGF-A-mediated angiogenesis in endometriotic lesion development resulting in a reduction in lesion number and size.

COX-2-mediated mechanisms also influence cellular proliferation and apoptosis in ectopic endometrium. PGE2 induced endometrial glandular epithelial cell proliferation in vitro (Jabbour and Boddy, 2003Go). Overexpression of COX-2 is associated with malignant proliferation (for a review see Gasparini et al., 2003Go) and is seen in endometrial cancer (Tong et al., 2000Go; Jabbour et al., 2001Go). Conversely epidemiological studies have shown that long-term use of non-selective NSAIDS reduces several tumour types in humans (Thun et al., 1991Go; Schreinemachers and Everson, 1994Go). Cell lines that overexpress COX-2 were resistant to apoptosis and had increased amounts of the survival factor Bcl-2 (Tsujii and DuBois, 1995Go). Bcl-2 has been identified at higher levels in ectopic endometrium when compared with eutopic endometrium from the same women (Jones et al., 1998Go). Additionally, eutopic endometrium from women with endometriosis had higher Bcl-2 levels and lower numbers of apoptotic cells when compared with normal controls (Meresman et al., 2000Go). We proposed that COX-2 inhibitors promote apoptosis and inhibit glandular proliferation in endometriotic tissue leading to a reduction in the size and number of lesions.

We chose the nude mouse model of endometriosis to test the hypothesis that inhibitors of COX-2 would disrupt several cellular processes involved in ectopic endometrial lesion formation causing a reduction in lesion size and number. In this model, human eutopic endometrium is implanted s.c. into T-cell-deficient estrogen-supplemented nude mice. The resulting lesions are morphologically and histologically similar to human disease (Bruner-Tran et al., 2002Go).

The selective COX-2 inhibitor nimesulide was used to ascertain if COX-2 inhibition reduces the size or number of ectopic endometrial lesions in the nude mouse model of endometriosis. Nimesulide inhibits COX-2 activity by 50% (IC50) at a 10-fold lower dose (1 µmol/l) than the IC50 for COX-1 (10 µmol/l) in human whole blood assays (FitzGerald and Patrono, 2001Go). The dose used in this study was consistent with the maximum tolerated dose used in an in vivo rat model of endometriosis (Gupta et al., 1999Go). A pharmacokinetic profile for nimesulide in the nude mouse model was established to optimize the dosage interval. Two independent experiments were performed using the nude mouse model of endometriosis to determine the influence of nimesulide on ectopic endometrial lesion number and size.


    Subjects and methods
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Ethical approval for the conduct of this study was obtained from the Cambridge Local Research Ethics Committee. All animal procedures were fully licensed under the Animal (Scientific Procedures) Act 1986 by the Home Office of the UK.

Recruitment of volunteers
Two female recruits aged 28 and 35 years gave informed consent to participate in this study. Both women had regular menstrual cycles (25 and 28 days), were not taking medications and were undergoing a therapeutic operative procedure for American Fertility Society stage III–IV endometriosis. Eutopic endometrium was obtained by pipelle suction curettage (Endocell, Wallach Surgical Devices Inc, Orange, CT) while the volunteers were anaesthetized. The secretory phase endometrium was collected on day 16 and 27, respectively, of the participant's menstrual cycle. The cycle phase of both biopsies was histologically confirmed using Noyes criteria (Noyes et al., 1950Go).

Endometrial biopsies were immediately placed in pre-warmed phenol-red free Dulbecco's modified Eagle's medium (DMEM)/F12 culture medium (Sigma-Aldrich Co. Ltd, Poole, UK).

Explant cultures
This method has been described previously (Bruner et al., 1997Go). Endometrial biopsies were washed with DMEM/F12 serum-free medium and sectioned into 2 mm3 fragments. Groups of 18 endometrial fragments were separated within tissue culture inserts. Each explant culture was incubated at 37°C in a humidified chamber with 95% air/5% CO2 and maintained in serum-free conditions in DMEM/F12 medium supplemented with 1% ITS (Invitrogen, Paisley, Scotland), 1% penicillin-G (Sigma-Aldrich Co. Ltd.), 0.1% Excyte (Serologicals Proteins Ltd., Kanakakee, IL) and 10 nmol/l estradiol (Sigma-Aldrich Co. Ltd) for 1–4 h.

Preparation of nude mice
Athymic female nude mice were purchased from Charles River, Manston, UK. The animals were housed in individually ventilated cages. The animal room was maintained at between 19 and 23°C with a 12 h light/12 h dark cycle, and mice were provided with a standard diet and water ad libitum. All procedures were performed in a laminar flow hood using sterile instruments. At 6 weeks of age, the mice were administered 0.10–0.15 ml of an anaesthetic mixture of Hypnorm [fentanyl citrate (0.045 mg/ml) and fluanisone (1.43 mg/ml)] (Roche, Welwyn Garden City, UK) and midazolam (0.7 mg/ml; Jassen Pharmaceutica, Beerse, Belgium) s.c. before performing bilateral ovarectomies. At the same time, estradiol-releasing pellets (1.5 mg/60 day release) (Innovative Research, Sarasota, FL) were implanted s.c. in the right flank of the nude mice. There was a minimum 10 day post-operative recovery period.

In vivo injection of human endometrial tissue
Human endometrial tissue was washed with sterile serum-free DMEM/F12. Each nude mouse received a single injection of 18 endometrial fragments in 150 µl of serum-free DMEM/F12 after numbing the skin with ethylchloride (BDH Laboratory Supplies, Poole, UK). S.c. injections were performed at a ventral midline site using 1 ml syringes and 19 gauge needles.

Experiment 1
Sixteen mice were assigned randomly to either the nimesulide-treated or control group. One mouse in the nimesulide group had to be excluded from the study as an infection developed at the site of the estrogen pellet insertion although it had developed an ectopic endometrial lesion.

Nimesulide (Sigma-Aldrich) was dissolved in polyethylene glycol (PEG) 300 (Sigma-Aldrich) in 25% normal saline to provide a dose of 25 mg/kg/day in a total volume of 0.1 ml. PEG 300 in 25% normal saline was the pharmacological control. Nimesulide and vehicle was injected s.c. into excess skin at the scruff of each animal's neck. The first dose was administered immediately after s.c. implantation of human endometrium. Subsequent doses were at 24 h intervals for 14 days.

Lesions were measured in a non-blinded fashion with callipers through the hairless nude mouse skin every 3 days. Fourteen days after endometrial tissue injection (day 14), mice were sacrificed between 0 and 20 h after the final dose of nimesulide or vehicle at specified time intervals. A pharmacokinetic profile for nimesulide was based on plasma samples taken at serial time intervals. Nimesulide plasma concentrations were determined by standard liquid chromatography and mass spectrometry techniques.

All mice developed at least one lesion. After dissection, all lesions were measured in three dimensions with callipers and a tissue volume estimate was calculated by multiplying the three dimension measurements together. If a mouse had more than one lesion, the volume estimate of each lesion was added together to give an estimate of total tissue load per mouse. Half the lesions were formalin fixed and the rest were snap-frozen and stored at –80°C. Haematoxylin and eosin staining was performed on all lesions.

Experiment 2
Fourteen mice were assigned randomly to receive nimesulide (n=7) or vehicle substance (n=7). On the basis of the results from experiment 1, three parameters were changed. Nimesulide was given at the same 25 mg/kg dose but at 12 h intervals to optimize the drug delivery according to the pharmacokinetic estimates. Twice daily drug administration was only continued for 10 days because lesion size differences were apparent by this stage in the first experiment. Finally, to avoid investigator bias, the final lesion measurements in experiment 2 were performed in triplicate in a blinded manner.

All lesions were measured and volume estimates calculated in triplicate as above. The mean of the triplicate volume estimates for each lesion was used.

Statistical analysis
In the first experiment, the small data set precluded an assumption of normal distribution, therefore the size of lesions in the nimesulide and control groups was compared on the basis of rank using the non-parametric Mann–Whitney U test. The probability that a new lesion in the nimesulide group would be smaller than a new lesion in the control groups was calculated using the formula [probability = Mann–Whitney U statistic/(n of group 1)x(n of group 2)].

In the second experiment, the mean of triplicate measurement values was used to rank the data. The Mann–Whitney U test was again used to compare the two groups. A power calculation was performed to assess the number of lesions required in each group to see a statistically significantly (P=0.05) difference in the means of 6 mm3 at a power of 80%.

Immunohistochemistry using formalin-fixed sections
Paraffin sections (5 µm) from nude mouse lesions were dewaxed and rehydrated followed by antigen retrieval with 0.1% trypsin (DIFCO Laboratories, Detroit, MI) in 0.1% CaCl2 (BDH Laboratory Supplies). An avidin–biotin blocking kit (Vector Laboratories Inc., Peterborough, UK) was used to reduce background staining.

COX-2 was detected using a monoclonal mouse antibody against human COX-2 (Zymed Laboratories Inc., San Francisco, CA). Mouse IgG (Sigma-Aldrich Co. Ltd) was used as the negative control. The mouse on mouse kit (Vector Laboratories Inc.) was used for non-specific background blocking and antibody detection.

After blocking with 20% goat serum (Sigma-Aldrich Co. Ltd) in 0.01% bovine serum albumin (BSA; Sigma-Aldrich Co. Ltd) in phosphate-buffered saline (PBS), mouse and human endothelial cells were identified using rabbit anti-von Willebrand factor (vWF) antibody (5.7 µg/ml) (Dako Corporation, Ely, UK). Non-specific rabbit immunoglobulins (5.7 µg/ml) (Dako Corporation) were used as a primary antibody control. Goat anti-rabbit immunoglobulin (7.5 µg/ml) (Vector Laboratories Inc.) and the Vectastain ABC system (Vector Laboratories Inc.) were used to detect the primary antibody.

A mouse monoclonal antibody against {alpha}-smooth muscle actin ({alpha}-SMA; Sigma-Aldrich Co. Ltd) (2.3 µg/ml) was used to identify myofibroblasts. An isotype-matched control (Sigma-Aldrich Co. Ltd) at the same concentration was used as a negative control. Non-specific background blocking and antibody detection was performed using the mouse on mouse reagent kit.

Murine F4/80-positive macrophages were identified in 5 µm acetone-fixed frozen sections. Non-specific antibody activity was blocked with 20% goat serum in 0.1% PBS/BSA. The sections were exposed to rat anti-mouse F4/80 (Serotec Inc., Raleigh, NC) (10 µg/ml) or an isotype-matched negative control (Serotec Inc.) (10 µg/ml) in 2% goat serum/0.1% BSA/PBS overnight at 4°C. The sections were washed in 0.1% Tween (BDH Laboratory Supplies) in PBS, then the secondary biotinylated goat anti-rat antibody (Zymed Laboratories Inc.) (1:4000) in 0.5% Tween/PBS was applied for 1 h at 37°C. After washing in PBS, the Vectastain ABC kit was used to detect antibody staining.

Diaminobenzidine (DAB; Sigma-Aldrich Co. Ltd) was used to visualize the antibody reactions. All sections were counterstained with Carazzi's haematoxylin, then rehydrated and mounted in Depex (BDH Laboratory Supplies). Microscopy was performed with an Ultraphot microscope (Carl Zeiss, New York, NY).


    Results
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Anti-human COX-2 antibodies detected COX-2 enzyme predominantly in the glandular epithelium of the ectopic endometrial lesions. Light staining was identified in the stromal compartment. Anti-COX-2 antibody was defined immunohistochemically by positive staining of the glandular epithelium of human eutopic endometrium (Figure 1), whereas minimal COX-2 staining was detected in mouse uterine tissue.



View larger version (178K):
[in this window]
[in a new window]
 
Figure 1. Immunohistochemical staining with anti-human COX-2 antibody. Anti-COX-2 antibody immunoreactivity in human glandular epithelium (A) but not mouse uterus (C). (B) and (D) are negative controls. The glandular epithelium of a nude mouse lesion exhibits positive COX-2 immunostaining (E). Background staining with a non-specific control antibody is minimal (F) (scale bar, A–F; 40 µm).

 
Nimesulide treatment did not influence lesion formation in either experiment. If the data from both experiments are combined, 14 of 14 nimesulide-treated mice and 14 of 15 control group mice developed lesions (Figures 2 and 4). The only mouse without a lesion was not exposed to nimesulide. Four mice in the nimesulide group and one in the control group developed more than one lesion (2–4) that were generally smaller than single lesions.



View larger version (9K):
[in this window]
[in a new window]
 
Figure 2. The three-dimensional size estimate (mm3) of ectopic endometrial deposits in nude mice exposed to 24 hourly doses of a control substance or nimesulide. The bar shows the average size of lesions in each group (mm3). The control group average size estimate were 15.9±6.4 mm3. The nimesulide group average size estimate was 14.1± 9.1 mm3.

 


View larger version (9K):
[in this window]
[in a new window]
 
Figure 4. The three-dimensional size estimate (mm3) of ectopic endometrial deposits in nude mice exposed to 12 hourly doses of a control substance or nimesulide. The bar shows the average size of lesions in each group (mm3). The control group average size estimate was 24.1±17.9 mm3. The nimesulide group average size estimate was 23.6±9.8 mm3.

 
The results of the first study showed no significant difference in volume estimate of tissue load between the nimesulide (total of 11 lesions from seven mice) and control groups (nine lesions from eight mice) (P=0.298, Mann–Whitney U test) (Figure 2). The mean volume estimate of lesions per mouse from the control group mice was 15.9±6.5 mm3, whereas the mean volume estimate of lesions per mouse from the nimesulide group was 14.1±9.1 mm3. The probability that a new observation from the nimesulide group would be smaller than a new observation from the control group was 64%. It was possible that a true effect of nimesulide was not detected because the drug dosage was inappropriate.

Although the nimesulide pharmacokinetic profile measurements were only made in one mouse at each time point, the profile was consistent with those seen in other in vivo models and in humans. The Cmax was 8200 nM, the Tmax was 1 h and the t1/2 was ~3.5 h (Figure 3). However, the free plasma concentration of nimesulide at 12 h was negligible. We therefore decided to perform a further experiment to determine if the trend towards smaller nimesulide lesions persisted when nimesulide was administered twice daily.



View larger version (19K):
[in this window]
[in a new window]
 
Figure 3. Pharmacokinetic profile of nimesulide when administered s.c. to nude mice at a dose of 25 mg/kg between 0 and 20 h post-dose.

 
When nimesulide was administered at 12 h intervals in experiment 2, the ranked difference in the volume estimate of tissue load between the nimesulide-treated (eight lesions from seven mice) and control groups (six lesions from seven mice) was not significant (P=0.53, Mann–Whitney U test) (Figure 4). The mean volume estimate of lesions per mouse in the control group was 24.1±17.9 mm3, whereas the mean volume estimate of lesions per mouse in the nimesulide group was 23.6±9.7 mm3. The probability that a new observation in the nimesulide group would be smaller than a lesion from the control group in this experiment was only 30%. A sample size calculation indicated that 138 mice in each group would be required to detect with 80% power a significant difference (P=0.05) in volume of 6 mm3 between the control and treated group means.

Nimesulide-treated and untreated lesions could not be differentiated histologically. All lesions contained glands lined with pseudostratified epithelium and central cystic spaces. Granulation tissue was identified peripherally, and small foci of necrosis and inflammation were observed centrally in implants from both the nimesulide and control groups (Figure 5A and B).



View larger version (142K):
[in this window]
[in a new window]
 
Figure 5. Minimal histological and immunohistochemical differences between control and nimesulide-treated nude mouse lesions. Haematoxyllin and eosin staining of a control (A) and treated (B) lesion. F4/80 antigen-positive murine macrophages in a control (C) and treated (D) lesion. Untreated (E) and nimesulide-exposed (F) nude mouse lesions containing {alpha}-smooth muscle actin-positive myofibroblasts. Blood vessels in control (G) and treated (H) lesions stained with von Willebrand factor (scale bar, A, B, E and F, 150 µm; C, D, G and H, 100 µm).

 
Myofibroblasts ({alpha}-SMA) (Figure 5C and D) and macrophages (F4/80) (Figure 5E and F) were identified in all lesions. No gross differences in macrophage and myofibroblast infiltration were seen between treatment groups.

An antibody to vWF was used to identify the vasculature supplying both nimesulide and control lesions. All lesions had a clearly identifiable vascular supply and there were no striking differences observed between lesions from nimesulide-treated and control mice (Figure 5G and H).


    Discussion
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
In this study, the COX-2 inhibitor nimesulide did not inhibit the formation or growth of ectopic human endometrial lesions in nude mice. These results suggest that COX-2 inhibitors are unlikely to influence the establishment or early progression of endometriosis in an estrogenic environment.

We explored several methodological reasons that could account for the negative results seen in this study. The human COX-2 enzyme was detected immunohistochemically in the glandular epithelium and to a lesser extent in the stroma of nude mouse lesions. This indicates that, like endometriosis, COX-2 is present in the relevant cells in this model system. Therefore, the nude mouse is suitable for investigation of the effect of nimesulide on ectopic endometrial growth.

Additionally, the dose of nimesulide used in the nude mouse model of endometriosis was sufficient to induce a pharmacological effect in rodents. When 25 mg/kg of nimesulide was administered s.c. to rats, carrageenin-induced paw inflammation was effectively reduced for the entire duration of the study (6 h) (Gupta et al., 1999Go). The serum levels of nimesulide that induced this anti-inflammatory effect were comparable with those seen in nude mice given the same 25 mg/kg dose of nimesulide. A much lower dose of nimesulide (5 mg/kg) was administered, 1 h prior to carrageenin injection in a rat pleurisy model of inflammation. A reduction in exudate volume and percentage of inflammatory cells was demonstrated in the rats that received nimesulide. PGE2 levels were also reduced in this treatment group (P<0.001), indicating that COX-2 inhibition was the mechanism of action of nimesulide's in vivo effect (Gilroy et al., 1998Go). Several other in vivo studies have also demonstrated a reduction in arachdonic acid metabolites after administration of nimesulide at lower doses than used in this study (for a review see Magni, 1993Go). Taken together, these studies indicate that the serum nimesulide levels that resulted from 12 hourly s.c. administration of 25 mg/kg of nimesulide were sufficient to inhibit COX-2 selectively in the nude mouse model of endometriosis.

The pharmacokinetic profile indicated that the maximum plasma concentration value (Cmax 18 mg/l) was achieved ~1 h (Tmax) after administration of 25 mg/kg/day nimesulide to nude mice. Healthy human volunteers given an oral dose of nimesulide (50–200 mg) tablets had pharmacokinetic profiles with a generally lower Cmax (1.9–9.85 mg/l) and a slightly longer Tmax (1.67–3.17 h) (Davis and Brogden, 1994Go). This difference is likely to reflect the s.c. administration and high dose chosen to ensure adequate nimesulide exposure in the nude mouse model. In our study, the t1/2 was 2–3 h, which is consistent with human profiles (1.9–4.5 h) (Davis and Brogden, 1994Go). Although the pharmacokinetic profile was only based on one plasma estimate at each time point, the results indicate that the dose of nimesulide administered to the nude mice was adequate to inhibit the COX-2 enzyme in vivo.

The whole blood assay is an in vitro measure of COX-2 selectivity. Thromboxane B2 production during blood coagulation reflects platelet COX-1 activity, whereas PGE2 production after lipopolysaccharide stimulation of whole blood is an index of macrophage COX-2 function (Brideau et al., 1996Go). The effect of nimesulide on these two parameters has been measured and the IC50 for COX-2 is 1 µmol/l, whereas the IC50 for COX-1 is 10 µmol/l in human whole blood assays (mean of four values from different studies) (FitzGerald and Patrono, 2001Go). The free plasma concentration of nimesulide in serum from nude mice is likely to have selectively inhibited COX-2 for almost the entire 12 h dosing interval. However, it is possible that COX-2 activity recovered at trough plasma concentrations and COX-1 inhibition occurred at peak plasma concentrations, just before and after nimesulide administration. Other highly selective COX-2 inhibitors may be more effective than nimesulide in reducing ectopic endometrial lesions in nude mice. However, the results of this study suggest that COX-2 inhibitors with a COX-2 selectivity profile like that of nimesulide (FitzGerald and Patrono, 2001Go) are unlikely to inhibit endometriotic lesion development.

Lesion formation and development occurred in the nude mouse despite the administration of a COX- 2 inhibitor at a therapeutic dose. Researchers using a mouse ear inflammatory assay have shown that COX-2 knockout mice maintain an inflammatory response to arachidonic acid (Morham et al., 1995Go). One explanation is that prostaglandins produced by homeostatic COX-1 activity contribute to inflammation. In our study, COX-1 activity may have been sufficient to permit lesion development in the nude mouse model.

It is possible that prostaglandin production by COX-1-mediated pathways in macrophages and fibroblasts may have masked nimesulide's therapeutic influence. In macrophages and fibroblasts that constitutively express COX-1, in vitro induction of COX-2 caused only a nominal 1.5- to 2-fold increase in PGH2 biosynthetic activity (Lee et al., 1992Go; Evett et al., 1993Go). Elevated levels of COX-1 mRNA have been detected in peritoneal macrophages harvested from women with severe endometriosis when compared with normal controls (Wu et al., 2002Go). We identified macrophages and myofibroblasts in both treated and untreated nude mouse lesions. Prostaglandin production in these cells may have been minimally affected by COX-2 inhibition due to high constitutive COX-1 activity.

Vascular disruption has been shown to influence lesion number and size in the nude mouse model of endometriosis (Hull et al., 2003Go). Nimesulide treatment did not alter these parameters, suggesting that products of COX-2 do not alter vessel formation. It is possible that the outcome measure of lesion size would not detect subtle changes in vessel development after nimesulide exposure. However, immunohistochemical identification of endothelial cells indicated that there were no differences in vascularity between treatment groups. It is therefore unlikely that COX-2 significantly influences ectopic endometrial vascular supply in the nude mouse model.

Any influence of nimesulide on cellular proliferation and apoptosis may not have been apparent as treatment was only for a 10 day period. Murine models evaluating nimesulide's effect on fast-growing lung tumour cell growth were able to determine growth inhibition after 7 days (Sheng et al., 1997Go; Williams et al., 2001Go). In the nude model of endometriosis, a longer time frame may be required to detect inhibition of the slower growth of non-invasive endometrial cells. As nude mice recover T-cell-mediated rejection of foreign tissue after 12 weeks of age, it is not feasible to assess such possible long-term effects in this model of endometriosis.

Eutopic endometrium from women with endometriosis in the secretory phase of the cycle was used to model the endometriotic disease process, as theorized by Sampson, as closely as possible in nude mice. Some researchers have argued that progesterone exposure in the secretory phase of the cycle inhibits lesion formation in the nude mouse (Bruner et al., 1997Go). Others have found that ectopic endometrial lesion development was independent of cycle phase in SCID mice (Aoki et al., 1994Go). In this study, the large numbers of lesions that developed in nude mice after injection of secretory phase endometrium (day 16 and 23) support the assertion that endometrial cycle phase does not influence nude mouse lesion development.

The development of endometriosis is dependent upon an estrogenic environment (Bulun et al., 2002Go), and 17B-aromatase (Noble et al.,1996Go) is elevated in endometriotic lesions. A possible positive feedback exists between aromatase activity and prostaglandins because PGE2 upregulates aromatase activity and estradiol stimulates COX-2 and PGE2 production (Noble et al., 1997Go; Bulun et al., 2000Go). COX-2 inhibitors are likely to reduce the COX-2-induced aromatase activity that has been identified in ectopic endometrial lesions. At a local level, this inhibition may occur in human ectopic endometrium. In nude mice, however, the effect of this inhibition may be obscured by iatrogenically administered estrogen.

The results of this study suggest that treatment with COX-2 inhibitors such as nimesulide is unlikely to reduce the size or number of endometriotic lesions in women, if administered in a constant estrogenic environment. In a study of histological samples obtained from 130 women (Fagotti et al., 2004Go), COX-2 immunostaining was more common in ovarian endometrioma (78.5%) than in peritoneal implants (11%) and rectovaginal nodules (13.3%). It may be that COX-2 inhibitors prove to influence ovarian endometriosis more than disease at other sites. Intraperitoneal lesions may also have a greater response to COX-2 inhibitors than nude mouse lesions that are supplied by peritoneal blood vessels (Hull et al., 2003Go) but lie outside the peritoneal cavity.

COX-2 inhibitors may also prove beneficial in the treatment of endometriosis-related pain unrelated to disease progression. Symptomatic improvement has been described in women with dysmenorrhoea (Pulkkinen, 1993Go) and women with endometriosis-related pelvic pain and dyspareunia (Corbellis et al., 2004Go) after treatment with COX-2 selective agents. Although COX-2 inhibitors have been shown to reduce gastrointestinal side effects significantly when used to treat rheumatoid arthritis, clinical trials have shown little improvement in disease progression when compared with non-selective COX inhibitors (Deeks et al., 2002Go).

In the nude mouse model of endometriosis, the COX-2 inhibitor nimesulide did not reduce lesion size or number, nor did it influence vascular development or macrophage and myofibroblast infiltration. Although COX-2 inhibitors may provide pain relief with fewer side effects than non-selective NSAIDs, the results of our study suggest that in an estrogenic environment, COX-2 inhibitors are unlikely to reduce ectopic endometrial tissue load in women.


    Acknowledgements
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
We would like to acknowledge Dr Brian D.M.Tom from the MRC Biostatistics Unit, Institute of Public Health, Cambridge for his advise on the statistical methods used in this study, and the patients and staff at The Rosie Hospital, Cambridge involved in this study. This study was funded by a research grant from Pfizer Global Research and Development, UK.


    References
 Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Aoki D, Katsuki Y, Shimizu A, Kakinuma C and Nozawa S (1994) Successful heterotransplantation of human endometrium in SCID mice. Obstet Gynecol 83, 220–228.[Abstract]

Ben-Av P, Crofford LJ, Wilder L and Hla T (1995) Induction of vascular endothelial growth factor expression in synovial fibroblasts by prostaglandin E interleukin-1: a potential mechanism for inflammatory angiogenesis. FEBS Lett 372, 83–87.[CrossRef][ISI][Medline]

Brideau C, Kargman S, Liu S, Dallob AL, Ehrich EW, Rodger W and Chan CC (1996) A human whole blood assay for clinical evaluation of biochemical efficacy of cyclooxygenase inhibitors. Inflamm Res 45, 68–74.[ISI][Medline]

Bruner KL, Matrisian LM, Rodgers WH, Gorstein F and Osteen KG (1997) Suppression of matrix metalloproteinases inhibits establishment of ectopic lesions by human endometrium in nude mice. J Clin Invest 99, 2851–2857.[Abstract/Free Full Text]

Bruner-Tran KL, Webster-Clair D and Osteen KG (2002) Experimental endometriosis: the nude mouse as a xenographic host. Ann NY Acad Sci 955, 328–339.[Abstract/Free Full Text]

Bulun SE, Zeitoun KM, Takayama K and Sasano H (2000) Estrogen biosynthesis in endometriosis: molecular basis and clinical relevance. J Mol Endocrinol 25, 35–42.[Abstract/Free Full Text]

Bulun SE, Yang S, Fang Z, Gurates B, Tamura M and Sebastian S (2002) Estrogen production and metabolism in endometriosis. Ann NY Acad Sci 955, 75–85.[Abstract/Free Full Text]

Chishima F, Hayakawa S, Sugita K, Kinukawa N, Aleemuzzaman S, Nemoto N, Yamamoto T and Honda M (2002) Increased expression of cyclooxygenase-2 in local lesions of endometriosis patients. Am J Reprod Immunol 48, 50–56.[CrossRef][ISI][Medline]

Corbellis L, Razzi S, De Simone S, Sartini A, Fava A, Danero S, Gioffre W, Mazzini M and Petraglia F (2004) The treatment with a COX-2 specific inhibitor is effective in the management of pain related to endometriosis. Eur J Obstet Gynecol Reprod Biol 116, 100–102.[CrossRef][ISI][Medline]

Daniel TO, Liu H, Morrow JD, Crews BC and Marnett LJ (1999) Thromboxane A2 is a mediator of cyclooxygenase-2-dependent endothelial migration and angiogenesis. Cancer Res 59, 4574–4577.[Abstract/Free Full Text]

Dannenberg AJ, Altorki NK, Boyle JO, Dang C, Howe LR, Weksler BB and Subbaramaiah K (2001) Cyclo-oxygenase 2: a pharmacological target for the prevention of cancer. Lancet Oncol 2, 544–551.[CrossRef][Medline]

Davis R and Brogden RN (1994) Nimesulide. An update of its pharmacodynamic and pharmacokinetic properties, and therapeutic efficacy. Drugs 48, 431–454.

Deeks JJ, Smith LA and Bradley MD (2002) Efficacy, tolerability, and upper gastrointestinal safety of celecoxib for treatment of osteoarthritis and rheumatoid arthritis: systematic review of randomised controlled trials. Br Med J 325, 619.[Abstract/Free Full Text]

Diaz-Flores L, Gutierrez R, Valladares F, Varela H and Perez M (1994) Intense vascular sprouting from rat femoral vein induced by prostaglandins E1 and E2. Anat Rec 238, 68–76.[ISI][Medline]

Dormond O, Foletti A, Paroz C and Ruegg C (2001) NSAIDs inhibit alpha V beta 3 integrin-mediated and Cdc42/Rac-dependent endothelial-cell spreading, migration and angiogenesis. Nat Med 7, 1041–1047.[CrossRef][ISI][Medline]

Evett GE, Xie W, Chipman JG, Robertson DL and Simmons DL (1993) Prostaglandin G/H synthase isoenzyme 2 expression in fibroblasts: regulation by dexamethasone, mitogens, and oncogenes. Arch Biochem Biophys 306, 169–177.[CrossRef][ISI][Medline]

Fagotti A, Ferrandina G, Fanfani F, Legge F, Lauriola L, Gessi M, Castelli P, Barbieri F, Minelli L and Scambia G (2004) Analysis of cyclooxygenase-2 (COX-2) expression in different sites of endometriosis and correlation with clinico-pathological parameters. Hum Reprod 19, 393–397.[Abstract/Free Full Text]

FitzGerald GA and Patrono C (2001) The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 345, 433–442.[Free Full Text]

Form DW and Auerbach R (1983) PGE2 and angiogenesis. Proc Soc Exp Biol Med 172, 214–218.[Abstract]

Gallo O, Masini E, Bianchi B, Bruschini L, Paglierani M and Franchi A (2002) Prognostic significance of cyclooxygenase-2 pathway and angiogenesis in head and neck squamous cell carcinoma. Hum Pathol 33, 708–714.[CrossRef][ISI][Medline]

Gasparini G, Longo R, Sarmiento R and Morabito A (2003) Inhibitors of cyclo-oxygenase 2: a new class of anticancer agents? Lancet Oncol 4, 605–615.[CrossRef][ISI][Medline]

Gilroy DW, Tomlinson A and Willoughby DA (1998) Differential effects of inhibitors of cyclooxygenase (cyclooxygenase 1 and cyclooxygenase 2) in acute inflammation. Eur J Pharmacol 355, 211–217.[CrossRef][ISI][Medline]

Gupta SK, Bhardwaj RK, Tyagi P, Sengupta S and Velpandian T (1999) Anti-inflammatory activity and pharmacokinetic profile of a new parenteral formulation of nimesulide. Pharmacol Res 39, 137–141.[CrossRef][ISI][Medline]

Haney AF and Weinberg JB (1988) Reduction of the intraperitoneal inflammation associated with endometriosis by treatment with medroxyprogesterone acetate. Am J Obstet Gynecol 159, 450–454.[ISI][Medline]

Hull ML, Charnock-Jones DS, Chan CL, Bruner-Tran KL, Osteen KG, Tom BD, Fan TP and Smith SK (2003) Antiangiogenic agents are effective inhibitors of endometriosis. J Clin Endocrinol Metab 88, 2889–2899.[Abstract/Free Full Text]

Jabbour HN and Boddy SC (2003) Prostaglandin E2 induces proliferation of glandular epithelial cells of the human endometrium via extracellular regulated kinase 1/2-mediated pathway. J Clin Endocrinol Metab 88, 4481–4487.[Abstract/Free Full Text]

Jabbour HN, Milne SA, Williams AR, Anderson RA and Boddy SC (2001) Expression of COX-2 and PGE synthase and synthesis of PGE2 in endometrial adenocarcinoma: a possible autocrine/paracrine regulation of neoplastic cell function via EP2/EP4 receptors. Br J Cancer 85, 1023–1031.[CrossRef][ISI][Medline]

Jones RK, Searle RF and Bulmer JN (1998) Apoptosis and bcl-2 expression in normal human endometrium, endometriosis and adenomyosis. Hum Reprod 13, 3498–3502.

Karck U, Reister F, Schafer W, Zahradnik HP and Breckwoldt M (1996) PGE2 and PGF2 alpha release by human peritoneal macrophages in endometriosis. Prostaglandins 51, 49–60.[CrossRef][Medline]

Keenan JA, Chen TT, Chadwell NL, Torry DS and Caudle MR (1995) IL-1 beta, TNF-alpha, and IL-2 in peritoneal fluid and macrophage-conditioned media of women with endometriosis. Am J Reprod Immunol 34, 381–385.[ISI][Medline]

Lee SH, Soyoola E, Chanmugam P, Hart S, Sun W, Zhong H, Liou S, Simmons D and Hwang D (1992) Selective expression of mitogen-inducible cyclooxygenase in macrophages stimulated with lipopolysaccharide. J Biol Chem 267, 25934–25938.[Abstract/Free Full Text]

Magni E (1993) The effect of nimesulide on prostaniod formation. Drugs 46, 10–14.[Medline]

Mahutte NG and Arici A (2003) Medical management of endometriosis-associated pain. Obstet Gynecol Clin North Am 30, 133–150.[ISI][Medline]

Masferrer JL, Leahy KM, Koki AT, Zweifel BS, Settle SL, Woerner BW, Edwards DA, Flickinger AG, Moore RJ and Seibert K (2000) Antiangiogenic and antitumor activities of cyclooxygenase-2 inhibitors. Cancer Res 60, 1306–1311.[Abstract/Free Full Text]

Matsumoto H, Ma WG, Daikoku T, Zhao X, Paria BC, Das SK, Trzaskos JM and Dey SK (2002) Cyclooxygenase-2 differentially directs uterine angiogenesis during implantation in mice. J Biol Chem 277, 29260–29267.[Abstract/Free Full Text]

Meresman GF, Vighi S, Buquet RA, Contreras-Ortiz O, Tesone M and Rumi LS (2000) Apoptosis and expression of Bcl-2 and Bax in eutopic endometrium from women with endometriosis. Fertil Steril 74, 760–766.[CrossRef][ISI][Medline]

Milne SA and Jabbour HN (2003) Prostaglandin-F2 alpha receptor expression and signalling in human endometrium: role of PGF2 alpha in epithelial cell proliferation. J Clin Endocrinol Metab 88, 1825–1832.[Abstract/Free Full Text]

Morham SG, Langenbach R, Loftin CD, Tiano HF, Vouloumanos N, Jennette JC, Mahler JF, Kluckman KD, Ledford A, Lee CA et al. (1995) Prostaglandin synthase 2 gene disruption causes severe renal pathology in the mouse. Cell 83, 473–482.[ISI][Medline]

Mori H, Sawairi M, Nakagawa M, Itoh N, Wada K and Tamaya T (1991) Peritoneal fluid interleukin-1 beta and tumor necrosis factor in patients with benign gynecologic disease. Am J Reprod Immunol 26, 62–67.[ISI][Medline]

Nie D, Lamberti M, Zacharek A, Li L, Szekeres K, Tang K, Chen Y and Honn KV (2000) Thromboxane A2 regulation of endothelial cell migration, angiogenesis, and tumor metastasis. Biochem Biophys Res Commun 267, 245–251.[CrossRef][ISI][Medline]

Noble LS, Simpson ER, Johns A and Bulun SE (1996) Aromatase expression in endometriosis. J Clin Endocrinol Metab 81, 174–179.[Abstract]

Noble LS, Takayama K, Zeitoun KM, Putman JM, Johns DA, Hinshelwood MM, Agarwal VR, Zhao Y, Carr BR and Bulun SE (1997) Prostaglandin E2 stimulates aromatase expression in endometriosis-derived stromal cells. J Clin Endocrinol Metab 82, 600–606.[Abstract/Free Full Text]

Noyes RW, Hertig AT and Roc KJ (1950) Dating the endometrial biopsy. Fertil Steril 1, 3–25.[ISI][Medline]

Ota H, Igarashi S, Sasaki M and Tanaka T (2001) Distribution of cyclooxygenase-2 in eutopic and ectopic endometrium in endometriosis and adenomyosis. Hum Reprod 16, 561–566.[Abstract/Free Full Text]

Pulkkinen M (1993) Nimesulide in dysmenorrhoea. Drugs 46 (Suppl 1), 129–133.[Medline]

Raiter-Tenenbaum A, Baranao RI, Etchepareborda JJ, Meresman GF and Rumi LS (1998) Functional and phenotypic alterations in peritoneal macrophages from patients with early and advanced endometriosis. Arch Gynecol Obstet 261, 147–157.[CrossRef][ISI][Medline]

Sakamoto Y, Harada T, Horie S, Iba Y, Taniguchi F, Yoshida S, Iwabe T and Terakawa N (2003) Tumor necrosis factor-alpha-induced interleukin-8 (IL-8) expression in endometriotic stromal cells, probably through nuclear factor-kappa B activation: gonadotropin-releasing hormone agonist treatment reduced IL-8 expression. J Clin Endocrinol Metab 88, 730–735.[Abstract/Free Full Text]

Schreinemachers DM and Everson RB (1994) Aspirin use and lung, colon, and breast cancer incidence in a prospective study. Epidemiology 5, 138–146.[ISI][Medline]

Sheng H, Shao J, Kirkland SC, Isakson P, Coffey RJ, Morrow J, Beauchamp RD and DuBois RN (1997) Inhibition of human colon cancer cell growth by selective inhibition of cyclooxygenase-2. J Clin Invest 99, 2254–2259.[Abstract/Free Full Text]

Smith WL, Garavito RM and DeWitt DL (1996) Prostaglandin endoperoxide H synthases (cyclooxygenases)-1 and -2. J Biol Chem 271, 33157–33160.[Free Full Text]

Strathy JH, Molgaard CA, Coulam CB and Melton LJ, 3rd (1982) Endometriosis and infertility: a laparoscopic study of endometriosis among fertile and infertile women. Fertil Steril 38, 667–672.[ISI][Medline]

Tamura M, Sebastian S, Yang S, Gurates B, Fang Z and Bulun SE (2002a) Interleukin-1{beta} elevates cyclooxygenase-2 protein level and enzyme activity via increasing its mRNA stability in human endometrial stromal cells: an effect mediated by extracellularly regulated kinases 1 and 2. J Clin Endocrinol Metab 87, 3263–3273.[Abstract/Free Full Text]

Tamura M, Sebastian S, Yang S, Gurates B, Ferrer K, Sasano H, Okamura K and Bulun SE (2002b) Up-regulation of cyclooxygenase-2 expression and prostaglandin synthesis in endometrial stromal cells by malignant endometrial epithelial cells. A paracrine effect mediated by prostaglandin E2 and nuclear factor-kappa B. J Biol Chem 277, 26208–26216.[Abstract/Free Full Text]

Thun MJ, Namboodiri MM and Heath CW Jr (1991) Aspirin use and reduced risk of fatal cancer. N Engl J Med 325, 1593–1596.[Abstract]

Tong BJ, Tan J, Tajeda L, Das SK, Chapman JA, DuBois RN and Dey SK (2000) Heightened expression of cyclooxygenase-2 and peroxisome proliferator-activated receptor-delta in human endometrial adenocarcinoma. Neoplasia 2, 483–490.[CrossRef][ISI][Medline]

Tsujii M and DuBois RN (1995) Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell 83, 493–501.[ISI][Medline]

Tsujii M, Kawano S, Tsuji S, Sawaoka H, Hori M and DuBois RN (1998) Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell 93, 705–716.[ISI][Medline]

Vane JR, Mitchell JA, Appleton I, Tomlinson A, Bishop-Bailey D, Croxtall J and Willoughby DA (1994) Inducible isoforms of cyclooxygenase and nitric-oxide synthase in inflammation. Proc Natl Acad Sci USA 91, 2046–2050.[Abstract/Free Full Text]

Williams CS, Sheng H, Brockman JA, Armandla R, Shao J, Washington MK, Elkahloun AG and DuBois RN (2001) A cyclooxygenase-2 inhibitor (SC-58125) blocks growth of established human colon cancer xenografts. Neoplasia 3, 428–436.[CrossRef][ISI][Medline]

Wu MH, Sun HS, Lin CC, Hsiao KY, Chuang PC, Pan HA and Tsai SJ (2002) Distinct mechanisms regulate cyclooxygenase-1 and -2 in peritoneal macrophages of women with and without endometriosis. Mol Hum Reprod 8, 1103–1110.[Abstract/Free Full Text]