Plasma Protein Binding of Warfarinin Healthy Volunteers

Br J clin Pharmac 1995; 39: 29S-33S

Effect of tenidap sodium on the pharmacodynamics and plasma protein binding of warfarin in healthy volunteers

G. APSELOFF', K. D. WILNER2 & N. GERBER' 'Ohio State University, College of Medicine, Department of Pharmacology, 5084 Graves Hall, 333 West Tenth Avenue, Columbus, OH 43210-1239 and 2Central Research Division, Pfizer Inc., Eastern Point Road, Groton, CT 06340, USA

1 An open-label, randomised study was performed to assess the effect of tenidap sodium on the pharmacodynamics and plasma protein binding of warfarin. 2 Fourteen healthy male volunteers received either a single oral dose of 120 mg tenidap sodium or matching placebo capsules from days 11 to 36. A single oral dose of 0.75 mg kg-1 warfarin was administered on days 1 and 32. 3 The mean prothrombin AUC(0,120h) value between baseline and day 32 increased from 1692.4 ± 234.5 s h to 1769.3 ± 218.0 s h in the group given tenidap, and decreased from 1747.6 ± 289.4 s h to 1708.1 ± 236.8 s h in the placebo group. 4 Tenidap caused a slight delay in the normalisation of prothrombin times following the second dose of warfarin on day 32 compared with the first dose on day 1. This was significant at 36, 48, 72 and 96 h but not at 120 h after administration of warfarin. 5 The mean percentage of unbound warfarin in the tenidap group (0.08% ± 0.09) was significantly different (P = 0.047) from that in the placebo group (-0.03% 0.10) but this was not considered to be clinically meaningful. 6 These data indicate that prothrombin times should be monitored during concomitant administration of tenidap and warfarin.

Keywords tenidap sodium warfarin pharmacodynamics plasma protein binding

Introduction Tenidap sodium is a novel, cytokine-modulating anti- Non-steroidal anti-inflammatory drugs (NSAIDs) rheumatic agent currently undergoing clinical investi- also inhibit cyclo-oxygenase; these include salicylates gation in the USA and Europe for the treatment of [10], ibuprofen [11-13], phenylbutazone [14] and rheumatoid arthritis and osteoarthritis. Tenidap inhibits flurbiprofen [15, 16], which have all been shown to cyclo-oxygenase activity [1] and has significant effects interact with warfarin, and these interactions may be on acute phase proteins; for example, tenidap has been of clinical importance [14]. Interactions may occur shown to lower the plasma concentration of C-reactive through a number of different mechanisms, including protein in patients with rheumatoid arthritis [2, 3]. displacement of warfarin (which is highly plasma pro- In vitro, tenidap has also been shown to inhibit protein bound) from plasma protein binding sites, and duction of the cytokines tumour necrosis factor, inter- stereoselective inhibition of the metabolism of the S- leukin-1 and interleukin-6 [4], which mediate release isomer of warfarin [14, 17]. The result is a prolonged of acute phase proteins from the liver [5]. prothrombin time without a proportional increase in Warfarin is a coumarin anticoagulant widely used in the concentration of warfarin. the treatment and prophylaxis of a variety of thrombo- More than 99% of tenidap is tightly bound to embolic disorders. It is used in the prophylaxis of plasma protein [18]; therefore, the effect of its admin- atrial fibrillation [6], thromboembolic complications istration to patients taking other highly bound drugs, with prosthetic heart valves [7], transient ischaemic such as warfarin, merits investigation. The potential attacks [8] and prophylaxis and treatment of deep clinical importance of an interaction between tenidap venous thrombosis [9]. and warfarin makes this study and assessment of the

Correspondence: Dr G. Apseloff, Ohio State University, College of Medicine, Department of Pharmacology, 5084 Graves Hall, 333 West Tenth Avenue, Columbus, OH 43210-1239, USA

© 1995 Blackwell Science Ltd 29S 30S G. Apseloff K. D. Wilner & N. Gerber effect of tenidap on prothrombin time as well as particles, preceded by a 50 x 3.9 mm guard column plasma protein binding in warfarin-treated subjects consisting of 40 ,um glass beads. The mobile phase both necessary and important. comprised 0.025 M Tris-phosphate buffer (pH 6.5) and methanol in a ratio of 55:45 (v/v) with a flow rate of 1 ml min-'. Absorbance was monitored at 365 nm Methods [20]. Standard curves, constructed over a concentration range of 0.5-25.0 Rg ml-', were run after every Subjects 30 samples. Mean correlation coefficients were 0.997 ± 0.002, and the mean slope was 0.379 ± 0.130. Healthy male volunteers who provided written, in- formed consent were enrolled in an open-label, Determination ofplasma protein binding of warfarin randomised, parallel-group study. At their initial screening visit all subjects underwent a medical Blood samples sufficient to provide 5 ml plasma history, full physical examination, urine drug screen, were collected immediately before the administration urinalysis, a 12-lead resting electrocardiogram, and of warfarin on days 1 and 32. Plasma was prepared standard laboratory tests for haematological, renal and and stored at -15° C until assayed for plasma protein hepatic functions, which were required to be within binding of warfarin by equilibrium dialysis. [ 14C]- 10% of normal limits for the subject to be eligible for Warfarin (46 jiCi mmol-'; Amersham) was added to a entry. Subjects were included only if they had not used 2.0 ml aliquot of plasma to a final concentration of 2.0 any prescription or over-the-counter medications in jg ml-' and 0.3 jiCi ml-' and was dialysed against 0.8 the previous 2 weeks, if they had a negative urine drug ml 0.13 M sodium phosphate buffer (pH 7.4) for 18 h screen and negative ethanol breath test, and were not at room temperature. Concentrations of warfarin in vegetarians. plasma and dialysate were measured by liquid scintillation counting using Ecolite scintillation fluid and a Protocol Packard scintillation counter. Subjects received a single oral dose of 0.75 mg kg-' Statistical evaluation warfarin (Du Pont Pharmaceuticals) on day 1 and either a once daily oral dose of 120 mg tenidap sodium The area under the prothrombin time-time curve during (Pfizer Central Research) (eight subjects) or matching the 120 h after administration of warfarin (AUC(0,120h)) placebo (six subjects) between days 11 and 36 inclu- was calculated by the linear trapezoidal method. Com- sive. On day 32, 2 h after the dose of tenidap or parison of the difference in AUC(0,120h) between the placebo, they received a second dose of 0.75 mg kg-l first and second doses of warfarin was undertaken warfarin. using a paired, two-tailed, Student's t-test. Mean dif- The protocol was approved by the Ohio State Uni- ferences in prothrombin time between the first and versity Human Subjects Biomedical Sciences Review second doses of warfarin were compared at each time Committee. Subjects were confined to the Research point (0-120 h). From scintillation counting data the Unit for at least 72 h after each dose of warfarin. percentage of unbound warfarin was calculated: Prothrombin times % unbound warfarin = 100 - [(dpm ml-' plasma - dpm ml-' buffer)/ Blood samples (3 ml) were obtained for determination dpm ml-' plasma]. of prothrombin times immediately before and at 12, The mean change in plasma protein binding of war- 24, 36, 48, 60, 72, 96 and 120 h after each dose of farin was also calculated using a paired, two-tailed warfarin. Prothrombin times were determined by a Student's t-test. standardised assay with thromboplastin and photo- optical detection using a Coag-A-Mate X2 (General Diagnostics) [19]. Results Determination ofplasma tenidap concentrations A total of 15 volunteers were enrolled into the study Blood samples sufficient to give 3 ml plasma were col- and their baseline demographic details are shown in lected prior to dosing with warfarin on day 1 and prior Table 1. Of these, 12 subjects completed the study and to dosing with tenidap or placebo on days 32 and 33. were included in the pharmacodynamic analysis. The Plasma was prepared and stored frozen at -15° C until three subjects who were withdrawn were all from the being used for assay. tenidap group: two were excluded for non-compliance In preparation for assay, acetonitrile (0.1 ml) con- (one before receiving tenidap), and the third subject taining CP-66,993 as an internal standard was added was withdrawn on day 34 (2 days after the second to 50 jil of plasma. The samples were vortexed and dose of warfarin) because of probable treatment- centrifuged at 40 C, and 50 jil of the supematant related adverse events (moderate nausea, vomiting and was added to vials containing 50 jil of 0.025 N Tris- abdominal discomfort). One subject in the placebo phosphate buffer (pH 7.4). Part of this sample (20 jil) group continued in the study despite developing skin was analysed by h.p.l.c. (Waters 481) using a 200 x bruising and myalgia approximately 1 week after the 4.6-mm Novopak C-18 column packed with 5 gm first dose of warfarin.

Table 1 Demographic characteristics of healthy male placebo groups (P < 0.05) (Table 2). The change in volunteers entered into a study to determine the effect of mean prothrombin times after the second and first dose 120 mg day-' tenidap vs matching placebo on prothrombin of warfarin for the tenidap and placebo groups was not time during concomitant treatment with 0.75 mg kg-l warfarin significantly different at 120 h. Warfarin + tenidap Warfarin + placebo Plasma protein binding of warfarin Number of subjects (all male) 8 6 The mean percentages of unbound warfarin determined from plasma obtained before the first and Age (years) second doses of warfarin (2 h after the dose of tenidap Range 22-45 20-36 Mean 33.0 25.8 on day 32) for subjects in the tenidap and placebo groups are shown in Table 3. The mean change of Weight (kg) 0.08% (range -0.06% to 0.18%) between day 32 and Range 64.6-84.1 70.4-92.7 Mean 73.5 77.2 day 1 for subjects in the tenidap group was significantly different (P = 0.047) from the mean change of Ethnic origin -0.03% (-0.15% to 0.10%) for subjects in the placebo White 8 5 Asian 0 1 group.

Plasma concentrations of tenidap Discussion Concentrations of tenidap in plasma were in the range of those observed previously in normal volunteers Concomitant administration of warfarin and 120 mg receiving daily doses of 120 mg tenidap (trough levels day-' of tenidap, compared with warfarin and placebo, ranged from 3.4 to 7.0 gg ml-') and indicated com- resulted in a small but statistically significant increase pliance with the study regimen [20]. in the prothrombin time AUC(0,120h). This is likely, however, to be of little or no clinical significance. A Prothrombin times slight delay in normalisation of prothrombin times in the tenidap group compared with placebo followed the All prothrombin times were within the normal range second dose of warfarin. The daily administration of (9.5-12.5 s) before administration of each dose of 120 mg tenidap had a small, but probably clinically warfarin. insignificant, effect on the plasma protein binding of Normalisation of mean prothrombin AUC(0,120h) values following the first and second doses of war- Table 3 Mean percentage of unbound warfarin in plasma farin are shown in Table 2. The mean difference in following administration of either 120 mg day-] tenidap or AUC(0,120h) between the first and second doses of matching placebo in warfarin-treated healthy volunteers warfarin in the tenidap group was significantly differ- = The Tenidap + warfarin Placebo + warfarin ent from that in the placebo group (P 0.007). (n =6) mean change in prothrombin time between the first (n =7) and second doses of warfarin was 4.5% in the tenidap Day I 1.00±0.09 1.04±0.07 group compared with -2.3% in the placebo group. Day 32 1.08 0.10 1.01 0.09 The differences in prothrombin times at 36, 48, 72 Day 32-Day 1 0.08 + 0.09* -0.04 ± 0.10 and 96 h between the second and first doses of warfarin were significantly different for the tenidap and *P = 0.047 vs placebo.

Table 2 Mean prothrombin time (s) and AUC(0, 120h) following administration of either 120 mg day-l tenidap or matching placebo to warfarin-treated healthy volunteers

Tenidap + warfarin (n = 5) Placebo + warfarin (n = 6) Time after Day 32 Day 32 warfarin dose (h) Day I Day 32 -day I Day I Day 32 -day I P value

0 10.60 10.94 0.34 10.75 10.92 0.17 NS 12 11.38 11.54 0.16 11.92 12.25 0.33 NS 24 15.30 15.74 0.44 15.73 15.78 0.05 NS 36 17.94 19.00 1.06 18.92 17.96 -0.70 0.02 48 17.60 18.60 1.00 18.77 17.18 -1.59 0.001 60 15.84 16.64 0.80 16.63 15.82 -0.81 0.08 72 14.58 15.02 0.44 15.05 14.20 -0.85 0.01 96 12.30 12.98 0.68 12.97 12.67 -0.03 0.04 120 11.20 11.96 0.76 11.82 11.76 -0.04 0.05 (NS) AUC(0,120h) (s h) 1692.4 1769.3 76.9 1747.6 1708.1 -39.5 0.007 ± 234.5 ±218.0 ±36.8 ±289.4* +236.8* +62.0*

*n = 5.

© 1995 Blackwell Science Ltd, British Journal of Clinical Pharmacology, 39, 29S-33S 32S G. Apseloff K. D. Wilner & N. Gerber warfarin. In the tenidap group the percentage of times after 1 week (P < 0.05). In four patients, bleed- unbound warfarin in plasma prior to the first dose ing times were prolonged above the normal range of compared with prior to the second dose of warfarin 140-570 s, and one of these patients experienced a increased by 0.08%. as opposed to a decrease of doubling of bleeding time. However, analysis of the 0.03% in the placebo group. results from this study was complicated by the many Warfarin is metabolised by the cytochrome P450- concomitant drugs some patients received. IIIA isozyme [21]. The lack of a clinically significant Drug interactions with antirheumatic medications interaction between tenidap and warfarin that is in- are a significant problem because of the numerous dicated by the present study suggests that tenidap medications utilised in the geriatric population. neither induces nor inhibits the P450111A isozyme. Elderly patients may be at further risk from decreased Similarly tenidap is unlikely to interact with other metabolic activity and reduced renal function. Aware- drugs which are metabolised by P450IIIA. ness of potential drug interactions is essential, and The effect of tenidap on plasma protein binding, appropriate monitoring should be undertaken to mini- although statistically significant, was small. Other anti- mise clinical toxicity. inflammatory agents, such as the NSAIDs mefenamic The findings of this study suggest that concurrent acid, phenylbutazone and salicylates, have also been treatment of warfarin and tenidap was safe and that the shown to displace warfarin from plasma protein bind- pharmacodynamics and plasma protein binding of ing sites [22]. However, clinically significant inter- warfarin were not altered to a clinically meaningful actions between different NSAIDs occur by extent by administration of tenidap. The statistically mechanisms other than displacement of plasma pro- significant increase in free warfarin in tenidap-treated tein binding, and the likelihood of clinical interactions subjects is also unlikely to be clinically relevant, and cannot be predicted on the basis of plasma protein there appears to be no evidence to discourage the binding [22]. advisability of co-administration of warfarin and Most NSAIDs potentiate oral anticoagulants via a tenidap. However, since many cyclo-oxygenase inhibi- number of mechanisms [14]. Schulman & Henriksson tors alter platelet function and haemostasis, additional [23] reported that in 20 patients (the majority elderly) monitoring of prothrombin time is recommended stabilised on warfarin, addition of ibuprofen (600 mg when tenidap is added to the regime of patients three times daily) significantly prolonged bleeding stabilised on warfarin.

References

1 Carty TJ, Showell HJ, Loose LD, Kadin SB. Inhibition 11 Slattery JT, Levy G. Effect of ibuprofen on protein bind- of both 5-lipoxygenase (5-LO) and cyclooxygenase (CO) ing of warfarin in human serum. J pharm Sci 1977; 66: pathways of arachidonic acid metabolism by CP-66,248, 1060. a novel antiinflammatory compound. Arthritis Rheum 12 Longenecker GL, Swift IA, Bowen RJ, Beyers BJ, Shah 1988; 31 (suppl 4): S89. AK. Kinetics of ibuprofen effect on platelet and endo- 2 Loose LD, Littman BH, Sipe JD. Modulation of acute thelial prostanoid release. Clin Pharmac Ther 1985; 37: phase proteins by tenidap, a new antiiflammatory, anti- 343-348. rheumatic drug. Clin Res 1990; 38: 579A. 13 McIntyre BA, Philp RB, Inwood MJ. Effect of ibuprofen 3 Loose LD, Sipe JD, Bartle L. Modulation of acute phase on platelet function in normal subjects and haemophiliac proteins by tenidap. Br J Rheumatol 1992; 31 (suppl 2): patients. Clin Pharmac Ther 1978; 24: 616-621. 175. 14 Day RO, Graham GG, Chapman GD, Lee E. Anti- 4 Sipe JD, Bartle LM, Loose LD. Modification of pro- rheumatic drug interactions. Clin Rheum Dis 1984; 10: inflammatory cytokine production by the antirheumatic 251-275. agents tenidap and naproxen. A possible correlate with 15 Market GA. Interaction study between phenprocoumon clinical acute phase response. J Immunol 1992; 148: and flurbiprofen. Curr med Res Opin 1977; 5: 26-32. 480-484. 16 Stricker BH, Delhes JL. Interaction between flurbiprofen 5 Heinrich PC, Castell JV, Andus T. Interleukin-6 and the and coumarins. Br med J 1982; 285: 1139. acute phase response. Biochem J 1990; 265: 621-636. 17 Furst DE. Clinically important interactions of non- 6 Albers GW, Atwood JE, Hirsh J, Sherman DG, Hughes steroidal antiinflammatory drugs with other medications. RA, Connolly SJ. Stroke prevention in nonvalvular JRheumatol 1988; 15 (suppl 17): 58-62. atrial fibrillation. Ann Intern Med 1991; 115: 727-736. 18 Gardner MJ, Wilner KD, McMahon GF, Fouda HG. 7 Stein PD, Kantrowitz A. Antithrombotic therapy in The pharmacokinetics of tenidap following single and mechanical and biological prosthetic heart valves and multiple 120 mg doses to healthy, male volunteers. Cliii saphenous vein bypass grafts. Chest 1989; 95 (suppl): Pharmac Ther 1993; 53: 211. 1075-1175. 19 Triplett DA, Harris CS. Procedures for the coagulation 8 Mirsen TR, Hachinski VC. Transient ischaemic attacks laboratory. American Society of Clinical Pathologists and stroke. Can med Ass J 1988; 138: 1099-1105. 1981; 12-14. 9 Anonymous. Management of venous thromboembolism. 20 Wilner KD, Gardner MJ. Tenidap sodium does not alter Lancet 1988; i: 275-277. the clearance or plasma protein binding of tolbutamide 10 Verbeeck RK. Pharmacokinetic drug interactions with in healthy male volunteers. Br J c-lini Pharmac 1995; 39: nonsteroidal antiinflammatory drugs. Clin Pharmacokin 39S-42S. 1990; 19: 44-46. 21 Kaminsky LS, Dunbar DA, Wang PP, et (11. Human

i 1995 Blackwell Science Ltd, B-itish Jouri-nil of Cliniol Pharmacology. 39. 29S-33S Effects of tenidap on warfarin 33S

hepatic cytochrome P-450 composition as probed by in 23 Schulman S, Henriksson K. Interaction of ibuprofen and vitro microsomal metabolism of warfarin. Drug Metab warfarin on primary haemostasis. Br J Rheumatol 1989; Dispos 1984; 12: 470-477. 28: 46-49. 22 Brater DC. Drug-drug and drug-disease interactions with nonsteroidal anti-inflammatory drugs. Am J Med 1986; 80 (suppl IA): 62-77.

i 1995 Blackwell Science Ltd, British Journal of Clinical Pharmacology, 39, 29S-33S