It has been ascertained that these drugs act by means of an action mechanism which blocks the synthesis of prostaglandin through the inhibition of the enzyme Cyclooxygenase (COX), known also as Prostaglandin G/H Synthase (PGHS). Through this action mechanism, the drugs belonging to the class of non-steroidal anti-inflammatory drugs (NSAID) show, alongside the anti-inflammatory action, negative side effects on the gastrointestinal apparatus, because prostaglandin synthesis is involved in both processes.

For many years it was therefore believed that it was impossible to obtain non- steroidal anti-inflammatory drugs (NSAID) that were not characterised by significant side effects on the gastrointestinal system. Later, after having shown that the enzyme cyclooxygenase (COX), which catalyses the conversion of arachidonic acid to prostaglandin and thromboxane, exists in two isoforms, known as COX-1 and COX-2, which proved to be codified by two different genes, it was also ascertained that the form indicated as cyclooxygenase-1 (COX-1) is virtually expressed in a constitutive way in all tissues and is therefore indicated as"constitutive", while the form cyclooxygenase-2 (COX-2) is expressed in response to stimuli, for example

of an inflammatory nature, and is indicated as"induced", even though many exceptions are known to the generalisation given above.

In particular, the"constitutive"form COX-1 is involved in the activity of the digestive tract, while the"induced"COX-2 isoform is involved in the inflammatory process.

Therefore it clearly appears the necessity to develop new drugs which selectively inhibit the"induced"form COX-2, involved in the inflammatory process, and which drugs are not inhibitors, or at least not selective inhibitors, of only the"constitutive"form COX-1, thus avoiding even serious adverse side effects on the gastrointestinal system.

Selective COX-2 inhibitors therefore possess a significant pharmacological activity as non-steroidal anti-inflammatory drugs for the treatment of acute pain syndromes and chronic disorders of an inflammatory nature. For example, selective COX-2 inhibitor such as Rofecoxib, Celecoxib (currently available on the market with the name Celebrex@) and Valdecoxib (currently available on the market with the name Bextra@) are advantageously used for the treatment of rheumatoid arthritis, osteoarthritis, and for the treatment of acute pain associated, for example, with dental surgery and with primary dysmenorrhoea. Recently, Rofecoxib, trade name Vioxx@, was withdrawn from the market due to data deriving from a study carried out on patients treated, who showed an alarming incidence of myocardial infarct and ischemia in comparison with the non treated group. In fact, this product, while demonstrating a good anti-inflammatory activity and the absence of side effects on the gastrointestinal system, showed, on the basis of these initial studies, serious and significant side effects on the cardiovascular system. For this reason several studies have been undertaken to assess any side effects on the cardiovascular system of selective anti-inflammatory product COX-2 inhibitors.

It therefore appears clear that the use of derivatives known as COX-2 inhibitors involves a series of consistent and dangerous side effects in the cardiovascular field, with respect to what has been shown for the non-

steroidal anti-inflammatory drugs that have been in use since a long time.

On the other hand, these compounds remain highly efficacious drugs for the treatment of inflammatory states, especially of medium and severe extent, and it is therefore important to interpret correctly the most significant side effects on the cardiovascular system that have recently been revealed.

Also during studies of cyclooxygenase inhibitors, a significant role of the constitutive form COX-1 was demonstrated in the development of carcinogenic pathologies, in particular in the development of intestinal polyposis and in the onset of cutaneous and ovarian carcinomas. In parallel, it was also observed that COX-1 plays an important role in the onset, for example following surgery, of medium and strong pain. Moreover, it was seen that the administration of COX-1 inhibitors can be useful in the prevention and/or treatment of atherosclerosis. It is therefore important to select and identify specific cyclooxygenase inhibitors even of the "constitutive"COX-1 form.

OBJECTS OF THE INVENTION The object of the present invention is to provide a class of compounds inhibitors of cyclooxygenase (COX-1 and/or COX-2), in particular 3, 4-diary isoxazole derivatives, which show a pharmacological activity and which are therefore used in the treatment of inflammation, in the prevention and treatment of carcinomas, in particular intestinal, ovarian and cutaneous, and in the treatment of pain syndromes, in particular resulting from surgery.

Another object of the present invention is to provide a process for the preparation of a class of compounds inhibitors of cyclooxygenase (COX-1 and/or COX-2), in particular 3, 4-diary isoxazole derivatives, which is simple, which consists of a number of limited steps, and which is industrially applicable.

Yet another object of the present invention is to provide a pharmaceutical composition which comprises at least one cyclooxygenase inhibitor (COX-1 and/or COX-2), in particular a 3, 4-diary isoxazole derivative, which is effectively used in the treatment of inflammation, in the prevention and

treatment of carcinomas, in particular intestinal, ovarian and cutaneous, and in the treatment of pain, in particular resulting from surgery.

An object of the present invention is also to provide a process for determining the degree of potential cardiovascular toxicity of cyclooxygenase (COX) inhibitor compounds, in particular COX-1 and/or COX-2 and their pharmaceutical compositions.

Then another aim of the present invention is to provide a process for determining the toxicity of cyclooxygenase (COX) inhibitor compounds, in particular COX-1 and/or COX-2, which is effective both in subjects who have not undergone any treatment with said compounds (preventive purpose) and in subjects who have been subjected to treatment with said compounds. The determination of the toxicity of these compounds, obtained through the process according to the present invention, provides the doctor with an important indicator which, together with other assessments normally made, allows the selection of patients who could be at cardiovascular risk if treated with selective COX-2 inhibitors.

Another object of the present invention is to provide a process for determining the actual activity of cyclooxygenase (COX) inhibitor compounds, in particular (COX-1 and/or COX-2).

Yet another object of the present invention is to provide a diagnostic kit for determining"in vitro"the degree of potential toxicity and/or the actual activity of cyclooxygenase (COX) inhibitor compounds, in particular (COX-1 and/or COX-2) and/or pharmaceutical compositions which comprise said compounds, which is rapid, reliable and economic.

DESCRIPTION OF THE INVENTION These and yet other objects and respective advantages which will be better explained by the following description, are achieved by cyclooxygenase inhibitor compounds (COX-1 and/or COX-2), in particular 3, 4-diary isoxazole derivatives, having the following general formula (2):

where: R is a C1-C5 alkyl linear, branched, substituted, not substituted chain; a Ci-C5 alkene linear, branched, substituted, not substituted chain; a C1-C5 alkyne linear, branched, substituted, not substituted chain, Ri and R2 are aryl, substituted aryl (where the term"aryl"also comprises heterocyclic compounds derived from arenes by substitution of one or more groups (-C=) and/or (-CH=CH-) with trivalent or divalent heteroatoms so as to maintain the electronic system characteristic of aromatic systems), with Ri=R2orRi2.

Particularly : R is chosen as methyl (CH3) or ethyl (CH3CH2) ; Ri is chosen among the following substituents: - (a) phenyl - (b) mesityl - (c) 5-chloro-2-furyl - (d) 2,4, 6- trimethoxyphenyl - (e) 2,4, 6-trimethoxy-3-chlorophenyl ; R2 is chosen as phenyl or benzenesulphonamide (C6H4SO2NH2).

According to a preferential aspect, R is chosen as methyl (CH3) or ethyl (CH3CH2) ; Ri is chosen among the following substituents: - (a) phenyl - (c) 5-chloro-2-furyl R2 is chosen as phenyl.

Always according to the present invention, the preferred compounds are the following : - compound P10 where: R is chosen as methyl (CH3), Ri is chosen as

(a) phenyl, R2 is chosen as phenyl ; - compound P9 where: R is chosen as ethyl (CH3CH2), R1 is chosen as (a) phenyl, R2 is chosen as phenyl ; - compound P6 where: R is chosen as methyl (CH3), Ri is chosen as (c) 5-chloro-2-furyl, R2 is chosen as phenyl.

The 3, 4-diary isoxazole derivatives, with a general formula (2) according to the present invention, are generally prepared by reaction between arylnitryle oxides and free enolate ions (in turn obtained by metallation reaction of various alkyl methyl ketones with LDA in suitable conditions), followed by a dehydration/aromatization reaction and if necessary by further reactions of derivatization/modification of the derivatives thus obtained.

The present invention also provides a process for determining the degree of potential toxicity affecting the cardiovascular system in subjects with inflammation for whom is indicated a treatment with drugs containing compounds that inhibit cyclooxygenase-2 (COX-2), and if necessary compounds that inhibit both cyclooxygenase-2 and cyclooxygenase-1, which comprises a step of determining the possible presence of TLR4 genetic polymorphism in these subjects.

The invention also provides a process for assessing the degree of potential toxicity affecting the cardiovascular system in subjects with inflammation undergoing treatment with drugs containing compounds that inhibit cyclooxygenase-2 (COX-2), and if necessary compounds that inhibit both cyclooxygenase-2 and cyclooxygenase-1, which comprises a step of determining the urinary levels of 11-dehydro-thromboxane B2 in isolated urine samples.

The invention also concerns a diagnostic kit for determining"in vitro"the degree of potential toxicity and/or the actual activity of compounds that inhibit cyclooxygenase (COX), in particular COX-1 and/or COX-2 and/or pharmaceutical compositions that comprise said compounds. The diagnostic kit is based on the determination of the presence of TLR4 genetic polymorphism and/or on the determination of the level of 11-dehydro-

thromboxane B2 in subjects before and after treatment with said cyclooxygenase inhibitor compounds and/or with pharmaceutical compositions that comprise them, or even in subjects who have not undergone said treatment.

The present invention presents several advantages. As is better specified below, the new compounds according to the invention proved to be highly selective ; the synthesis of this class of compounds according to the invention allows excellent yields to be obtained; the process of determining TLR4 polymorphism in subjects to be treated with anti-inflammatory drugs makes it possible to assess whether COX-2 inhibitors may or may not give rise to phenomena of significant toxicity affecting their cardiovascular apparatus.

Moreover, that determination allows the assessment of the actual activity of the COX-1 and COX-2 inhibitor compounds.

The synthetic procedure for the preparation of 3, 4-diaryl isoxazole derivatives, of general formula (2) is indicated in SCHEME 1 shown below : SCHEME 1 With reference to the scheme shown above, the key passage concerns the preparation of the desired enolate (21) from the corresponding ketone (20).

Theoretically, the ketone may lead to the formation of two different enolates in the presence of a base. According to the present invention, in the indicated synthetic step, conditions of precise thermodynamic control are used, in particular the reaction with LDA is carried out at the temperature of 0°C, so as to generate selectively the desired enolate. Later, the enolate is

reacted with various arylnitryle oxides (22) to give the 5-hydroxy-2- isoxazoline compounds (1). These compounds are obtained with good yields and with good diastereoisomeric ratios, as shown in Table 1.

Table 1. Yields and diastereoisomeric ratios for the reaction between (21) and (22).

Compound R1 ~ drb la- (a) phenyl 61 85 : 15 1b- (b) mesityl 29 70: 30 1c- (c) 5-chloro-2-furyl 45 80: 20 1 d- (d) 2,4, 6-trimethoxyphenyl 52 70: 30 1e (e) 2,4, Strimethoxyßchlorphenyl 60 70: 30 a) the yields refer to recrystallized products. b) the diastereoisomeric ratios were determined by means of H NMR.

The diastereoisomeric ratios were determined by'H NMR spectroscopic analysis of the reaction crudes, in which the major isoxazoline isomer 1 (cis) has a more shielded Sens than that of the minor isomer (trans), as was ascertained in analogous systems. In fact, the chemical shift 5cH3 of the major isomer, for example the cis 1a, is closer (1.27 ppm) to that of an already known compound, cis-3, 5-dimethyl-4-phenyl-5-hydroxy-2-isoxazoline (C), than it is (to the same already known compound) that of the minor isomer trans. On the contrary, the 8CH3 of the minor isomer (trans 1a) is closer to that of an already known compound trans-3-methyl-4-phenyl-5- hydroxy-2-isoxazoline (D). The effect of the phenyl group is more evident if one examines (B) where it is not present, and the chemical shift of methyl is 1.71 ppm, about the same as (D) in which methyl and phenyl are in trans.

The following SCHEME 2 shows a comparison between known compounds and compounds 1a trans and cis according to the present invention, used for H NMR spectra interpretation.

SCHEME 2 \/ //\ --oh \ CH3 N\ H3 ;", H Ha (A) SOH < 3 3 bcH3=1. 1 ppm bcH3=1. 71 ppm 85-CH3=1. 21 ppm 8CH= 4. 00 ppm l \ l \ l \ H3 H3 9 Q ß Q X OV 0'CH3 cisl trans l a) 65-CH3= 1. 69 ppm aCH3= 1. 27 ppm aCH3= 1. 77 ppm 8CH= 4. 00 ppm aCH-4. 52 ppm bCH= 4. 45 ppm

The reaction between (22) when Ri is chosen as (b) mesityl or (c) 5-chloro-2- furyl and enolate ions of ketones (21), also leads to the formation of appreciable quantities of corresponding isoxazoles of formula (2), thanks to the basic conditions used during the reaction. All the above is simplified in Table 2.

Table 2. yields of isoxazolines (1b and 1c) and isoxazoles (2b and 2c) formed during the reaction between (22) when Ri is chosen as (b) mesityl or (c) 5-chloro-2-furyl and enolate ions of ketones (21).

Ri Reaction time yield (h) lsoxazoline Isoxazole Mesityl 3 29 (1b) 55 (2b) Mesityl overnight-- (1b) 73 (2b) 5-Chloro-2-furyl"45 (1c) 13 (2c) a) the Yields refer to products isolated by chromatography.

The 5-hydroxy isoxazoline compounds (1) according to the invention are then

converted into the corresponding isoxazole compounds (2) in basic conditions. The yields and the reaction conditions are shown below in Table 3.

Table 3. Yields of the isoxazoles (2) obtained from the compounds with formula (1).

Compound Ri Reaction yield time 2a- (a) phenyl 1 h 80 2b- (b) mesityl 15'50 2c- (c) 5-chloro-2-furyl 1 h 60 2d- (d) 2,4, 6-trimethoxyphenyl 2 h 74 2e (e) 2,4, 6-timethoxy-3-chlorophenyl 2 h 75 a) the yields refer to products isolated by chromatography.

Also according to the present invention, the compounds with general formula (2) obtained according to the scheme indicated above can be advantageously further derivatized/modified, for example they may be subjected to an alkylation reaction as shown in example 1 for the synthesis of the product P9 (3, 4-diphenyl-5-ethylisoxazole).

According to the above synthesis scheme, it was also possible to synthesise the compound Valdecoxib or 4- (5-methyl-3-phenylisoxazol-4- yl) benzenesulphonamide, a powerful and selective inhibitor of COX-2, obtaining an overall yield (55%) greater than that given in the literature (32%).

The compounds according to formula (2) can also be further subjected to various derivatization reactions, in particular of the substituents Ri and/or R2 which substitute the isoxazolic ring. The reaction of the compounds 2 with CIS03H/NH40H was particularly interesting, for example in the case of the compounds 2a, 2c and 2e as shown below in SCHEME 3: SCHEME 3

The yields are given below in Table 4: Compound Ri Yield 3a (a) phenyl 85 3c (c) 5-chloro-2-furyl 73 3e (e) 2,4, 6-trimethoxy-3-chlorphenyl 23 a) the yields refer to products isolated by chromatography.

In the case of the reaction of the product 2b with CISO3H/NH4OH, the products obtained 4b and 5b are indicated below in the synthesis SCHEME 4: SCHEME 4

The yields are given below in Table 5: Isoxazole/yield CIS03H 4b 5b 1 : 8 16 44 1: 11 60- a) the yields refer to products isolated by chromatography.

Instead, in the case of compound 2e, treatment with CIS03H/NH40H leads to the formation of the corresponding product 3e, together with 6e and 7e, deriving from the halogenation of 2e (SCHEME 5).

SCHEME 5 The products to which the present invention refers, and in particular the compounds P6, P9 and P10 proved to be excellent COX-1 inhibitors.

Given below are the chemical analyses of the above-mentioned compounds, obtained according to the general method described above.

EXAMPLE 1 Synthesis of the product P9 (3, 4-diphenyl-5-ethylisoxazole). n-Butyllithium in hexane (2.19M, 0.213 mL, 0.4675 mmol) was added to 3,4- diphenyl-5-methylisoxazole (0.100 g, 0.425 mmol) in THF (5 mL) kept under stirring at-78 °C under nitrogen, using a three-neck flask under a flow of nitrogen with magnetic stirring, an entry for nitrogen and a dropping funnel.

The red reaction mixture obtained was kept under stirring for 1 hour at-78 °C before adding CH31 (4.25 mmol). The reaction mixture was brought to room

temperature and then processed by adding NH4CI aqueous solution. The two phases were separated and the aqueous phase was extracted three times with ethyl acetate. The combined organic extracts were dried on anhydrous Na2SO4 and then the solvent was evaporated in a vacuum. The residue was subjected to column chromatography (silica gel, petroleum ether: ethyl acetate = 10/1) and 3, 4-diphenyl-5-ethylisoxazole was obtained with a yield of 75 %.

3, 4-Diphenyl-5-ethylisoxazole (2g). 75% yield. mp 85-87°C (hexane), white crystals. FT-IR (KBr): 3029,3005, 2923,2848, 1625,1596, 1493,1467, 1437,1410, 1327,1282, 1210,1011, 905,771, 702 cm~1. 1H NMR (200 MHz, CDCI3 d) : 1.29 (t, 3H), 2.78 (q, 2H); 7.12-7. 43 (m, 10H). 13C NMR (75 MHz, CDCI3 J) : 12.52, 19.67, 115.20, 127.90, 128. 67,128. 93,129. 41,129. 53, 130. 16, 130.66, 161.34, 171.43. GC-MS (70 eV) m/z (rel. int. ): 249 (M+, 100), 234 (6), 221 (18), 220 (99), 194 (9), 193 (61), 192 (46), 165 (17), 115 (7), 103 (8), 89 (53), 77 (15), 63 (10), 51 (10). Anal. calc for C17H15NO : C, 81.90 ; H, 6.06 ; N, 5.62. Found: C, 81.92 ; H, 6.02 ; N, 5.60.

EXAMPLE 2 General synthesis of the compounds 2 A solution of Na2CO3 (2.24 mmol) in water (10 mL) was added to 3-aryl-4- phenyl-5-hydroxy-5-methyl-2-isoxazoline (1.12 mmol) in THF (10 mL). The reaction mixture was then heated under reflux for the time indicated in table 3. The two phases were separated and the aqueous phase was extracted three times with ethyl acetate. The combined organic phases were dried on anhydrous Na2SO4 and the solvent was evaporated at reduced pressure to give the isoxazoles with a yield of 50-80%.

Analytical data of the product P6 (2c) 3- (5-Chloro-2-furyl)-4-phenyl-5- methylisoxazole (2c).

60% yield. mp 71-73°C, yellow crystals. FT-IR (KBr): 3147,3051, 2927, 2848, 1633, 1520, 1435, 1412, 1236, 1204, 1134, 1020, 985,940, 926,897, 796,775, 704 cm~1. 1H NMR (CDCl3, 8) : 2.36 (s, 3H); 6.11-6. 12 (d, 1H, J= 3.57 Hz); 6.25-6. 27 (d, 1 H, J= 3.57 Hz); 7.25-7. 30 (m, 2H, aromatic protons);

7.40-7. 47 (m, 3H, aromatic protons). 13C NMR (75 MHz, CD3, 6) : 11.41, 108. 14, 113. 87, 114. 99, 128. 58, 129. 01, 129. 63, 130. 20, 138. 59, 143. 76, 152.42, 167.10. GC-MS (70 eV) m/z (rel. int.) : 261 [M (37CI) +, 5], 259 [M (35CI) +, 15], 219 (11), 217 (36), 154 (17), 127 (10), 115 (5), 102 (5), 89 (14), 77 (9), 63 (10), 51 (12), 43 (100). Anal. calculated for C14H10NO2CI : C, 64.75 ; H, 3.88 ; N, 5.39. found: C, 64. 73; H, 3.85 ; N, 5.38.

EXAMPLE 3 Analytical data of the product P10 (2a) 3, 4-Diphenyl-5-methylisoxazole (2a).

80% yield. mp 97-98°C (hexane), white crystals. FT-IR (KBr): 3051, 2928, 1619,1597, 1573,1497, 1464,1436, 1414,1376, 1304,1239, 1074,915, 769,696 cm~1. 1H NMR (CDCl3, 8) : 2.45 (s, 3H); 7.17-7. 47 (m, 10H). 13C NMR (75 MHz, CDCl3, 8) : 11.82, 116.01, 127.90, 128.69, 128.95, 129.38, 129.60, 130.05, 130.61, 161.39, 166.85. GC-MS (70 eV) m/z (rel. int.) : 235 (M+, 100), 220 (28), 194 (14), 193 (90), 192 (37), 165 (28), 103 (10), 90 (12), 89 (62), 78 (10), 77 (24), 63 (23), 51 (48), 43 (70). Anal. calculated for C16H13NO : C, 81. 68 ; H, 5.57 ; N, 5.95. found: 81.73 ; H, 5.50 ; N, 5.95.

The products obtained according to the present invention were tested with an in vitro test on whole human blood isolated as described in Patrignani P. et al., J Pharmacol Exp Ther 1994 ; 271 : 1705-12 ; Patrignani P et al., J Clin Invest 1982 ; 69 : 1366-1372 ; Patrono C. et al.. Low dose aspirin and inhibition of thromboxane B2 production in healthy subjects. Thromb. Res. 1980 ; 17 : 317-27. The assays were carried out to assess the effects of the substituents and the activity of the compounds according to the present invention as inhibitors of COX-1/COX-2 activity, taking Valdecoxib (present trade name Extra0) as a reference.

11 healthy volunteers (6 females aged between 25 and 29 years) were enroled to participate in the study after each having given their written approval. The same volunteers had previously taken part in other studies.

COX-2 Assay 1 ml aliquots of a sample of isolated peripheral venous blood samples

containing 10 IU of sodium heparin were incubated in the laboratory in the presence of LPS (lipopolysaccharide) (10microg/ml) or saline for 24 hours at 37°C as described. The contribution of platelet COX-1 activity was suppressed by pre-treating the subjects with 300 mg of aspirin 48 hours before blood sampling. Plasma was separated by centrifugation (10 minutes at 2,000 r. p. m. ) and kept at-70°C until assayed for prostaglandin (PG) E2, as an index of LPS-induced monocyte COX-2 activity.

COX-1 Assay.

Samples of isolated peripheral venous blood were used in vitro (taken from the same donors when they had not taken any non-steroidal anti- inflammatory drug during the two weeks preceding the study). 1 ml aliquots of isolated whole blood were immediately transferred into glass test tubes and kept at a temperature of 37°C for 1 hour. The serum was separated by centrifugation (10 min at 3000 rpm) and kept at-70°C until assayed for thromboxane (TX) B2. (TX) B2 whole blood production was measured as a reflexion of the maximally stimulated platelet COX-1 activity in response to endogenous formed thrombin.

Effect of the tested compounds on isolated whole blood COX-1 and COX-2 activities.

The compounds (0.005-150 mM) were dissolved in DMSO and aliquots of 2 microlitres of the solutions were pipetted directly into test tubes to give final concentrations of 0.01-300 microM in the blood. From 4 to 9 different concentrations of each compound were incubated with heparinised isolated whole blood samples in the presence of LPS (10 micrograms/ml) for 24 hours or with isolated whole blood samples allowed to clot at 37°C for 1 hour, in order to examine the concentration-dependence of COX-2 vs. COX-1 inhibition, respectively.

Analysis of PIGEZ and TXB_ The concentrations of PGE2 and TXB2 were measured by radioimmunological assays (Patrono C. et al.. Low dose aspirin and inhibition of thromboxane B2 production in healthy subjects. Thromb. Res. 1980; 17:

317-27; Ciabattoni G. et al. J Endocrinol Invest 1979; 2: 173-182). Isolated plasma and serum samples were diluted in the standard diluent of the assay (0.02M phosphate buffer, pH 7.4) and assayed in a volume of 1.5 ml at a final dilution of 1: 50-1: 30.000. 4000 d. p. m. of [3H]-PGE2 or [3H]-TXB2 were used and specific anti-PGE2 and anti-TXB2 antibody diluted 1: 100,000 and 1: 120,000, respectively. The least detectable concentration was 1-2 pg/ml for both prostanoids.

Results LPS-stimulated isolated whole blood samples, drawn from healthy subjects treated with aspirin 300 mg 48 h before sampling, produced 335. 8 ng of PGE2 per ml of plasma (meanS. E. M., n=13). TXB2 production in clotting isolated whole blood samples, obtained from the same subjects in aspirin- free periods, averaged 46653 ng/ml (meanS. E. M. , n=13).

As shown in figures 1,2, 3, and 4, the compounds P9, P10, P6 and the reference Valdecoxib inhibited LPS-induced monocyte COX-2 and thrombin- stimulated platelet COX-1 activities in a concentration-dependent fashion.

IC50 values for inhibition of platelet COX-1 and monocyte COX-2 activities are given in Table 6.

Valdecoxib inhibits monocyte COX-2 and platelet COX-1 activities with the following values of IC5o : 27 (16.22-44. 63) and 0.57 (0.4619-0. 7056) microM (95% interval).

The analysis of the sigmoidal concentration-response curves for inhibition of monocyte COX-2 and platelet COX-1 by the compound P6, showed that virtually complete suppression (>90%) of platelet COX-1 activity occurs at concentrations that do not affect monocyte COX-2 activity (Figure 4). Similar results were obtained with the compounds P9 and P10 (Figure 1 e 2).

Table 6. Pharmacological data.

COX-1 COX-2 Compound R1 substituent R IC50(µM) IC50(µM)<BR> on R2

2a (P10) phenyl H CH3 0. 0 (0.05-0. 144) (1.577- 3.954) 2g (P9) phenyl H CH3CH2 0.05a 1.49c (0.028- (0.812- 0.097) 2.768) 2b mesityl H CH3 >100 >100 2c (P6) 5-chloro-2-furyl-H CH3 0. 5a >100a (0.2641- 0.9248) 2d 2,4, 6-trimethoxyphenyl H CH3 >100 >100 2e 2,4, 6-trimethoxy-3- H CH3 >100 >100 chlorophenyl Valdexocib phenyl SO2NH2 CH3 27d 0.57e (16.22- (0. 4619- 44.63) 0.7056) 5b 2,4, 6-trimethyl-3- H CH3 >100 >100 benzenesulphonamide 3c 5-chloro-2-furyl SO2NH2 CH3 >100 >100 3d 2,4, 6-trimethoxy-3-SO2NH2 CH3 >100 >100 chlorophenyl 4b 2,4, 6-trimethyl-3-SO2NH2 CH3 >100 >100 benzenesulphonamide The values are means of at least two experiments an=3 b n=5 cn=4 d n=11 en=13 (values in parentheses are the IC50 confidence intervals) The inventors surprisingly found a significant role of the receptor TLR4 (Toll-

like receptor 4) in the onset and progression of toxicity affecting the cardiovascular system of subjects treated with COX-2 inhibitors.

Recently, epidemiological and experimental studies have revealed an important role of TLR4 in the progression and development of atherosclerosis (Cook et al. Nature Immunology 2004; 5: 975-9). In particular, an association has been reported between the presence of the polymorphisms of TLR4 and a reduced atherosclerosis and also a lower risk of acute coronary events for the carriers of said polymorphism (Kiechl et al. New Engl J Med. 2002; 347: 185-9; Edfelt et al. Eur Heart J 2004; 25: 1447-53; Ameziane et al 2003; 23: 61-4).

Thromboxane A2 and prostacyclin which, as has already been said, represent the major products of the metabolism of arachidonic acid in the platelets and in the endothelial cells, perform a fundamental role in cardiovascular homeostasis. In particular, thromboxane A2 causes platelet aggregation, vasoconstriction and vascular proliferation, while prostacyclin inhibits the platelet aggregation induced by many agonists, the vascular proliferation of the smooth muscle cells and vascular tone. Thus while thromboxane A2 promotes the initiation and the progression of atherogenesis, prostacyclin prevents it. The side effects on the cardiovascular system, caused by the administration of cyclooxygenase-2 (COX-2) inhibitors, would therefore be due to the reduction of prostacyclin (which performs an athero-protective role), caused by the inhibition of COX-2 which participates in and contributes to the biosynthesis of prostacyclin in human. The enzyme COX-1, the principal isoform of COX responsible for the biosynthesis of thromboxane A2 (which has a negative effect on the vascular system), not being inhibited by the anti-inflammatory compounds that selectively inhibit COX-2, does not undergo any variations and maintains the standard levels of thromboxane A2. Thus, the balance between the athero-protective and anti-thrombotic effect of prostacyclin (the biosynthesis of which depends mainly on COX-2) and the"negative"effect of thromboxane A2 (the biosynthesis of which depends mainly on COX-1) is

altered by COX-2 selective inhibitors of in favour of thromboxane A2, with consequent reduction of the"protective"effect of prostacyclin, and increase of damage to the cardiovascular system.

The inventors assessed the biosynthesis of prostacyclin and of thromboxane in subjects carriers of TLR4 polymorphism (Asp299Gly and Thr39911e) in comparison with"wild-type" (WT) subjects who do not present said polymorphism, matched for age, sex and cardiovascular risks factors, in the absence of treatment with Aspirin@.

From the experimental results given below in detail, the following was surprisingly found.

In comparison with wild-type subjects, in subjects who were carriers of TLR4 polymorphisms, the urinary levels of 11-dehydro-TXB2 and of 2,3-dinor-6- keto-PGF1 a, systemic indices of thromboxane A2 and of prostacyclin biosynthesis respectively, were significantly reduced. Moreover the ratio between the biosynthesis of prostacyclin and of thromboxane A2 (prostacyclin/thromboxane ratio) was significantly higher in the carriers of TLR4 polymorphism than in wild-type subjects. Thus, in subjects who are carriers of TLR4 polymorphism, the reduced biosynthesis of platelet thromboxane seems to contribute to a greater cardiovascular protection and to reduced atherosclerosis, in comparison with wild-type subjects.

This consideration was used as the starting point for carrying out a study assessing the degree of potential toxicity of selective COX-2 inhibitors affecting the cardiovascular system.

Treatment with rofecoxib causes a similar inhibition of prostacyclin (about 60%) in the two groups of patients and is associated with an increased urinary excretion of 11-dehydro-TXB2 in carriers of TLR4 polymorphisms, but not in wild-type subjects. When treated with selective COX-2 inhibitors, wild- type subjects undergo a reduction of the biosynthesis of prostacylin (COX-2 dependent) with a consequent reduction of the"protective"activity with respect to the cardiovascular system, typically carried out by prostacyclin as already said. Since they are selective COX-2 inhibitors, COX-1 platelet

activity remains more or less unchanged, and so the level of thromboxane A2 in vivo remains unchanged, and the consequent"negative"effect on the cardiovascular system. The result is a moderate onset of risk factors for the cardiovascular system, given by the decrease of prostacyclin, without a significant reduction of the levels of thromboxane As.

When treated with selective COX-2 inhibitors, carriers of TLR4 polymorphisms also undergo a reduction of the biosynthesis of prostacylin (COX-2 dependent), but they show a contemporary and surprising increase in the production of thromboxane A2. This increase is so significant that the values of thromboxane A2 may reach, and even exceed, the levels normally present in the wild-type population and so they are such as to bring back to at least standard conditions the carriers of TLR4 polymorphisms (who, as has already been said, in the absence of treatment, have lower levels of Thromboxane A2 and therefore represent a"cardioprotected"phenotype with respect to WT subjects).

So the presence of TLR4 polymorphism seems to amplify dramatically the consequences of the inhibition of vascular prostacyclin by selective COX-2 inhibitors on platelet function.

In practice, in subjects who are carriers of TLR4 polymorphism treated with selective COX-2 inhibitors, the moderate effect of potential cardiovascular toxicity, which can be found in WT subjects for the reasons indicated above, is enormously amplified, thanks to the increase in the level of thromboxane A2 with respect to what is observed in the absence of treatment with COX-2 inhibitors in the same carriers of TLR4 polymorphism, and therefore results in a potential cardiovascular toxicity. So, while selective COX-2 inhibitors could be administered to WT subjects with a moderate and at any rate acceptable risk of complications for the cardiovascular system, the administration of the same compounds to TLR4 polymorphism subjects would involve an extremely high and probably unacceptable risk for damages to the cardiovascular apparatus. So for these subjects characterised, in the absence of treatment with COX-2 inhibitors, by a reduced urinary excretion of

11-dehydro-TXB2, the administration of selective COX-2 inhibitors would be unadvisable, but this would not be the case in WT subjects, for whom the beneficial effects of anti-inflammatory activity could be higher than the potential cardiovascular toxicity.

The results shown above were obtained on the basis of the following assessments.

Subjects studied.

The study protocol was approved by the Ethical Committee of the University of Chieti"G. d'Annunzio", Faculty of Medicine.

The written approval of all the subjects studied was obtained.

408 outpatients were examined who took part in a programme for assessing cardiovascular risk at the Department of Internal Medicine.

The genotype of all for TLR4 polymorphism Asp299GIy and Thr39911e was assessed, according to the method of Lorenz et al. (Biotechniques 2001; 31: 22-4). 26 subjects proved to be carriers of TLR4 polymorphism.

In short, the DNA extracted from the peripheral blood was amplified with PCR using two specific pairs of primer for the regions containing Asp299GIy and Thr39911e polymorphisms : Asp299G1y, forward 5'-gattagcatacttagactactacctccatg-3' ; reverse 5'- gatcaacttctgaaaaagcattcccac-3' ; Thr39911e, forward 5'ggttgctgttctcaaagtgattttgggagaa-3'; reverse 5'acctgaagactggagagtgagttaaatgct- The two PCR products were digested using restricting enzymes Nco I and Hinf I, respectively. After digestion, the samples were made slide on 3% agar gel. For Asp299Gly polymorphism, the wild type allele was represented by a band of 249 bp, while the mutant allele was shown by a band of 226 bp. For Thr39911e polimorfismo, the wild type allele was represented by a band of 406 bp, while the mutant allele was shown by a band of 377 bp.

An allele frequency of 2.8% was calculated and a carriage rate of 6.3%.

Seven subjects were excluded from the study because they were under treatment with aspirin for cardio-protection. The 19 selected subjects were subdivided as follows : 9 heterozygous for Asp299G1y and Thr39911e, 5 heterozygous for Asp299Gly and"wild-type"for Thr39911e, 3"wild-type"for Asp299G1y and heterozygous for Thr39911e, 1 heterozygous for Thr39911e and homozygous for Asp299G1y and 1"wild-type"for Asp299G1y and homozygous for Thr39911e. These 19 subjects were compared with 19 subjects who were not carriers of TLR4 polymorphism, matched for age, sex and cardiovascular risk factors. The characteristics of the subjects studied are listed below in Table 7.

Table 7. Baseline characteristics of subjects who were carriers of TLR4 polymorphisms Asp299G1y and Thr39911e and of wild-type subjects. <BR> <BR> <BR> <BR> <BR> <BR> <P> "wild-type"TLR4 TLR4 Polymorphisms<BR> <BR> <BR> <BR> <BR> <BR> No. 19 19 Age (years) (meanDS) 5611 5611 sex, F (%) 10 (53%) 10 (53%) Body mass index, (Kg/m2) 307 307 (meanDS) hypertension, n° (%) 14 (73) 15 (79) Diabetes, n° (%) 2 (10) 3 (16) Smokers, n° (%) 2 (10) 2 (10) Cholesterol LDL, mg/dl 13547 12327 (meanSD) Drugs Nitrates, n° (%) 0 (0) 1 (5) ß-Blockers, n° (%) 1 (5) 4 (21) Calcium channel blockers, 4 (21) 6 (32) n° (%) ACE inhibitors, n° (%) 13 (68) 9 (47)

Diuretics, n° (%) 3 (16) 5 (26) Statins, n° (%) 3 (16) 3 (16) All the subjects were studied by means of standard interviews, clinical examination and biochemical analyses. General biochemical assessments were carried out with traditional laboratory methods. Cholesterol LDL was calculated using the Friedewald formula. At the time of assessments, the subjects did not present any infection in progress or any inflammatory condition.

Study protocols.

For the 19 carriers of TLR4 polymorphism and for the 19 control carriers of "wild-type"TLR4, overnight samples (from 8 p. m. to 8 a. m. ) were collected to determine, on isolated samples, the levels of 11-dehydro thromboxane B2, one of the principal enzymatic metabolites of Thromboxane A2. 11-dehydro- thromboxane B2 represents an index of the systemic biosynthesis"in vivo"of Thromboxane A2. The urinary levels of 2, 3-dinor-6-keto-prostaglandinF1a were also determined, which is one of the principal enzymatic metabolites of prostacyclin and represents an index of its systemic biosynthesis"in vivo", and the values of 8-iso-prostaglandin F2a, index of oxidative stress"in vivo".

Samples of whole blood were also collected and isolated to assess"in vitro" platelet COX-1 activity, as determined by the measurement of serum Thromboxane B2, and of COX-2 monocyte activity, determined by the production of PGE2 on whole blood stimulated with lipopolysaccharide (Escherichia coli) (10 pg/ml) for 24 hours, as reported in Patrignani et al., J Pharmacol Exp ther 1994 ; 271: 1705-12.

Six subjects who were carriers of TLR4 polymorphism and 6"wild-type"TLR4 control subjects were treated with rofecoxib (a selective COX-2 inhibitor) (trade name VIOXXO) (25 mg a day, at 8 a. m. ) for 5 consecutive days.

Subjects with disturbances of a haemorrhagic nature, allergy to aspirin or to other non-steroidal anti-inflammatory drugs (NSAID) or with a history of gastrointestinal disorders were excluded from the study. The subjects

refrained from using aspirin or other non-steroidal anti-inflammatory drugs (NSAID) for at least two weeks prior to the study. The blood samples were taken to determine COX-1 platelet activity and COX-2 monocyte activity before treatment with the drug and 12 hours after the last administration (at 8 a. m. on the sixth day of the study).

The collection of overnight urine samples was obtained to assess the urinary excretion of 11-dehydro-thromboxane B2 and of 2, 3-dinor-6-keto-PGF1a before treatment with the drug and after the last dose of rofecoxib.

The study was completed before the voluntary withdrawal of Rofecoxib from the market by Merck due to the significant increase of the incidence of serious and significant side effects at cardiovascular level (Fitzgerald GA. N Engl J Med 2004 ; 351: 1709-11).

Biochemical analyses.

The levels of TXB2, PGE2, 11-dehydro-TXB2, 2, 3-dinor-6-keto-PGF1a and 8- iso-PGF2a were measured with previously validated radioimmunological techniques (Patrono C, Ciabattoni G, Pinca E, et al. Low dose aspirin and inhibition of thromboxane B2 production in healthy subjects. Thromb Res 1980 ; 17 : 317-27. Patrignani P, Panara MR, Greco A, et al. Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther 1994 ; 271 : 1705-12. Ciabattoni G, Maclouf J, Catella F, et al.

Radioimmunoassay of 11-dehydrothromboxane B2 in human plasma and urine. Biochim Biophys Acta 1987 ; 918 : 293-7. Wang Z, Ciabattoni G, Creminon C, et al. Immunological characterization of urinary 8-epi- prostaglandin F2 alpha excretion in man. J Pharmacol Exp Ther 1995 ; 275 : 94-100).

The levels of soluble CD40L, of the chemokine MCP-1 and of the vascular cellular adhesion molecule (VCAM-1) in citrated plasma were determined by ELISA determination-enzyme-linked immunosorbent assay (R & D System) according to the manufacturer's instructions. The concentration of fibrinogen in the plasma was determined by analysing its enzymatic conversion into

fibrin. The C-reactive protein in the serum (CRP) was determined with the nephelometric method (Behring).

Statistical analysis.

All the values are given as median and range, unless otherwise indicated.

The comparisons between the groups were made with the x2 test and the Mann-Whitney test. The correlations were determined with the Spearman rank-test. The statistical significance was indicated with P<0.05. In the pharmacological study, the primary hypothesis was that rofecoxib (25 mg/day) would cause a 60% reduction in the urinary excretion of 2,3-dinor-6- keto-PGF1a. Assuming an inter-subject coefficient of variation of 30% for the urinary excretion of 2, 3-dinor-6-keto-PGF1a, the primary end-point, in carriers of TLR4 polymorphisms, 6 volunteers should have allowed a difference of 62% to be found in its measurement with a power of 90% based on a two- tailed test with values of P lower than an error factor of 0.05 of the first type.

6 TLR4 polymorphism carrier subjects and 6 non carriers were therefore enrolled. The software SPSS Inc (version 11.5 for Windows) was used for all the statistical analyses.

The results will now be commented with reference to the enclosed figures, where: - fig. 5 shows the levels of biological markers of inflammation in carriers of TLR4 polymorphism and in"wild-type"TLR4 carriers. Levels of VCAM-1 (A), fibrinogen (B), CD40L (C), chemokine MCP-1 (D) and seric PCR (E) were assessed. The urinary excretion of 8-iso-prostaglandin F2a, an index of oxidative stress"in vivo", was assessed in the collection of night urine (from 8 p. m. to 8 a. m. ) (F). The data are reported as a mean and a confidence interval. [*P=0.008 for the comparison between the carriers of TLR4 polymorphism and of'Swild-type"TLR4].

- fig. 6 refers to the biosynthesis of prostacyclin and thromboxane A2 in carriers of TLR4 polymorphism and in"wild-type"TLR4 carriers. Box-and- Whisker plot of urinary excretion of 2, 3-dinor-6-keto-prostaglandin F1a (A),

one of the principal enzymatic metabolites of prostacyclin which is an index of its systematic biosynthesis in vivo, and 11-dehydro-thromboxane B2 (B), one of the principal enzymatic metabolites of thromboxane A2 which is an index of its systematic biosynthesis in vivo, and the prostacyclin/thromboxane A2 ratio (C), in overnight urine collection (from 8 p. m. to 8 a. m. ) of 19 subjects who were carriers of TLR4 polymorphism and 19"wild-type"controls. The data are reported as a mean and a confidence interval. (*P=0.041, **P<0.0001, P=0.0061 for the comparison between the carriers of TLR4 polymorphism and of'vvild-type"TLR4). Correlation between 2, 3-dinor-keto- prostaglandin Fla and 11-dehydro-thromboxane B2 in carriers and non carriers of TLR4 polymorphism (D).

- fig. 7 shows the effects of rofecoxib (25mg per day for 5 consecutive days) on the biosynthesis in vivo of prostacyclin and thromboxane in carriers of TLR4 polymorphism (A) and in carriers of"wild-type"TLR4 (B). The data are reported as means and standard errors. [*P= 0.0087, §P=0. 041, #P=0. 0043 with respect to the values prior to taking the drug (pre-drug)].

Levels of inflammatory biomarkers.

As shown in figure 5A-E, the circulating levels of plasma soluble VCAM-1T but not of fibrinogen, of MCP-1, CD40L and CRP, were significantly lower in carriers of TLR4 polymorphisms than in the wild-type [473 (324-681) versus 624 (260-1301) ng/ml, median (range), respectively, P=0.008]. Moreover, the formation of F2-isoprostanes was studied (determined by measuring the urinary excretion of 8-iso-PGF2a, a non-invasive index of oxidative stress in vivo) in the two groups of subjects. As shown in fig. 5F, comparable levels of urinary excretion of F2-isoprostane were found in carriers of TLR4 polymorphisms and in the wild-type.

Biosynthesis of prostaglandin E2 in whole blood stimulated with lipopolysaccharide.

It was determined whether the leukocytes of the carriers of TLR4 polymorphisms show a deficit of recognition of the lipopolysaccharide of

Escherichia coli, assessing the production of PGE2 in whole blood stimulated with lipopolysaccharide (cf. Patrignani P, Panara MR, Greco A, et al.

Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther 1994 ; 271: 1705-12). The production of PGE2 was not significantly different in the carriers of TLR4 polymorphisms and in the wild-type [22 (6- 78) versus 33 (12-174) ng/ml, respectively, P=0.328] (not shown).

Biosynthesis of prostacyclin and thromboxane in vivo.

The biosynthesis of prostacyclin and of thromboxane in vivo was determined by measuring the urinary levels of 2, 3-dinor-6-keto-PGF1 a and of 11- dehydro-TXB2, respectively. The urinary excretion of 2, 3-dinor-6-keto-PGF1 a and 11-dehydro-TXB2 was significantly lower in carriers of TLR4 polymorphisms than in wild-type subjects [122 (50-577) versus 188 (86-436) pg/mg creatinine, respectively, P=0.041] (Fig. 6A).

Likewise, the urinary excretion of 11-dehydro-TXB2 was significantly lower in carriers of TLR4 polymorphisms than in wild-type subjects [174 (63-462) versus 540 (211-836) pg/mg creatinine, respectively, P<0.0001] (Fig. 6B).

In order to exclude a defect of platelet COX-1 in carriers of TLR4 polymorphisms which could have resulted in a reduced biosynthesis of thromboxane A2 in vivo, we determined the production of TXB2 in whole blood allowed to clot. The levels of serum TXB2 did not present a statistically significant difference in the two groups [1.5 (0.6-2. 1) versus 1.6 (0.8-2. 9) ng/108 platelets, respectively, P=0. 157] (not shown).

The ratio of prostacyclin/thromboxane A2 biosynthesis was significantly higher in the carriers of TLR4 polymorphisms than in the wild-type subjects [0.8 (0.3-2. 2) versus 0.5 (0.11-1. 1), respectively, P=0.006] (Fig. 6C).

Moreover, a statistically significant correlation was found between the urinary excretion of 2, 3-dinor-6-keto-PGF1 a and 11-dehyro-TXB2 in carriers of TLR4 polymorphisms but not in wild-type subjects [rS=0. 665, n=38, P=0.002 versus rs=-0. 07547, n=38, P=0.7588, respectively] (Fig. 6D).

Effects of Rofecoxib (COX-2 inhibitor) on the biosynthesis of prostanoids.

Rofecoxib (V ! OXX@), a selective inhibitor of COX-2, was administered in a dose of 25 mg per day, for 5 consecutive days to 6 carriers and non carriers of the genetic polymorphism of TLR4. The effects of Rofecoxib on COX-1 platelet activity were assessed (assessing the production of serum TXB2 as described in Patrono et al., Thromb Res 1980 ; 17: 317-27) and on COX-2 monocyte activity (measuring the production of PGE2 in whole blood stimulated with LPS, as described in Patrignani et al., J Pharmacol Exp Ther 1994 ; 271: 1705-12). In carriers and non carriers of TLR4 polymorphism, the administration of Rofecoxib does not significantly influence COX-1 platelet activity [-11 (from 0 to-57) vs-10 (from 0 to-10) % of variation vs pre- treatment values, respectively, P=0.489], while it causes a profound inhibition of monocyte COX-2 activity ex vivo [-94 (from-77 to-98) vs-85 (from-69 to- 97) % inhibition, respectively, P=0.589] (Table 8). Rofecoxib causes a selective inhibition of the activity of cyclooxygenase-2 in the two groups of subjects.

Table 8. Effects of rofecoxib (25 mg per day) administered for 5 consecutive days to 6 carriers of TLR4 polymorphisms and 6 wild-type subjects on the biosynthesis of prostanoids ex vivo and in vivo. Pre-drug Rofecoxib "Wild-type" TLR4 Serum Thromboxane B2, ng/ml 252.5 (176-384) 233.5 (178-391) Prostaglandin E2 induced by 20.3 (14.3-58. 1) 2.3 (0.3-4. 2) lipopolysaccharide, ng/ml 11-dehydro-thromboxane B2 pg/mg 454 (227-819) 345 (155-595.6) creatinine 2,3 dinor-6 keto-prostag ! andin Fia, 174.5 (77-229) 56.5 (38-84) pg/mg creatinine Carriers of TLR4 Polymorphisms Serum Thromboxane B2, ng/ml 286.5 (125-617) 225 (94-728) Prostaglandin E2 induced by 45. 6 (22-67) 5.3 (1. 9-11) lipopolysaccharide, ng/ml 11-dehydro-thromboxane B2 pg/mg 150 (116-301) §-329 (158-587)' creatinine 2,3 dinor-6-keto-prostag) andin Fia, 117 (78-285) 50 (28-93)' pg/mg creatinine

** P<0.01, *P<0.05, rofecoxib with respect to the values prior to taking the drug; §P<0. 01, values prior to taking the drug in carriers of TLR4 polymorphisms with respect to"wild-type"subjects.

The effects of Rofecoxib on the biosynthesis of systemic prostacyclin and of Thromboxane A2 were evaluated. As shown in Figure 7 and in Table 8, Rofecoxib causes a similar inhibition of the urinary excretion of 2,3-dinor-6- keto PGF1o [-64 (from-42 to-72) vs to-61 (from-35 to-83) %, respectively, P=0.818] in the two groups, which was associated with an increased urinary excretion of 11-dehydro-TXB2 in carriers of TLR4 polymorphisms but not in the"wild-type"subjects [174 (136-354) vs 71 (67-84) % of pre-drug values, respectively, P=0.002] (Table 8).

As has already been said, in the absence of treatment with GOX-2 inhibitors, the subjects with TLR4 Asp299GIy and Thr39911e polymorphisms show a decrease of the onset of atherosclerosis and a reduced risk of acute coronary events. From the studies carried out, it was seen that the inhibition by Rofecoxib of the biosynthesis of COX-2-dependent prostacyclin (about 60%) is associated with an increased urinary excretion of 11-dehydro-TXB2 in carriers of genetic TLR4 polymorphisms but not in"wild-type"subjects, that is in non carriers of polymorphism. The above results lead us to conclude that TLR4 polymorphism increases the consequences of the inhibition of vascular prostacyclin on platelet function by the inhibitors of cyclooxygenase-2. The administration of cyclooxygenase-2 inhibitors in these subjects, carriers of TLR4 polymorphisms, transforms the cardiovascular protection phenotype into an appreciably exposed phenotype which can lead to acceleration of atherogenesis and to an exaggerated thrombotic response following the

rupture of the atherosclerotic plaque.

The above results therefore indicate the possibility of identifying and selecting the subjects (that is carriers of TLR4 polymorphisms) who may suffer severe damage from treatment with selective COX-2 inhibitors, with the same initial conditions of the subjects and treatment conditions. So, in these subjects who are carriers of TLR4 polymorphism, the treatment with anti-inflammatory cyclooxygenase-2 inhibitor drugs may provoke serious side effects in the cardiovascular field and thus give sever problems of toxicity.

In consideration of the above, the present invention therefore provides a process for determining the degree of potential toxicity of compounds that are at least inhibitors of cyclooxygenase-2 (COX-2), and possibly of compounds that inhibit both cyclooxygenase-2 and cyclooxygenase-1, which comprises a step of determining the possible presence of genetic TLR4 polymorphism in the subjects to be treated with drugs based on the above-mentioned COX-2 inhibitor compounds.

More in particular said polymorphism is a genetic TLR4 polymorphism (Asp299G1y and/or Thr39911e) which is determined"in vitro"by means of PCR amplification using two pairs of primers for the regions containing the Asp299GIy and Thr39911e polymorphisms.

The carriers of polymorphism are the subjects for whom the toxicity of said COX-2 inhibitor compounds is greater. This results in the use of compounds that inhibit at least cyclooxygenase-2 (COX-2) as anti-inflammatory drugs only for subjects without genetic TLR4 polymorphism (Asp299G1y and/or Thr39911e). Said compounds are chosen for example among: Rofecoxib (VIOXX@), Celecoxib (Celebrex@), Valdecoxib (Bextra@) and Lumiracoxib (Prexige@).

The invention also provides a process for assessing the degree of potential toxicity of compounds that inhibit at least cyclooxygenase-2 (COX-2) which comprises a step of determining the levels of thromboxane A2 in vivo by measuring the urinary levels of 11-dehydro-thromboxane B2. In other words, the levels of thromboxane A2 of the subject are indirectly determined by

analysing the level of 11-dehydro-thromboxane B2 on isolated urine samples.

As already mentioned, a previous selection of the subjects treated or to be treated based on the presence or absence of TLR4 polymorphism, as well as being critical for the assessment of the possible toxicity of the treatment with COX-2, will also allow the assessment of the activity of the drugs used in a more certain and coherent way. In other words, the invention provides a process for determining the actual activity of cyclooxygenase inhibitor compounds, in particular cyclooxygenase-1 (COX-1) and/or cyclooxygenase- 2 (COX-2), which comprises a step of determination"in vitro"of the genetic TLR4 polymorphism (Asp299G1y and/or Thr39911e) by means of PCR amplification using two pairs of primers for the regions containing the Asp299G1y and Thr39911e polymorphisms, and/or a step of determination of the levels of thromboxane A2 in vivo by measuring the urinary levels of 11- dehydro-thromboxane B2.