COST-EFFECTIVENESS OF VALDECOXIB COMPARED TO DICLOFENAC IN PATIENTS WITH RHEUMATOID ARTHRITIS (RA) IN THE UNITED KINGDOM (UK) AND GERMANY R. H. Alten, MD;1 B. von Scheele, MS, MSc Pharm;2 L. Gaffney, MSc;3 F. Maciver, MSc;3 L. Niculescu, MD, MPP, MPH4 1Schlossparkklinik Berlin, Germany; 2RTI Health Solutions, Södra Sandby, Sweden; 3RTI Health Solutions, Manchester, UK; 4Pfizer Inc, New York, NY, USA.

INTRODUCTION • Cost-effectiveness of valdecoxib and diclofenac was Figure 6. German cost-effectiveness acceptability curve for compared using country-specific unit costs for resource symptomatic ulcer avoided, valdecoxib 20 mg qd over 6 months. • Cyclooxygenase-2 (COX-2) specific inhibitors have use (hospital days, study medication, concomitant similar efficacy profiles to nonspecific nonsteroidal 100 medication, unscheduled procedures, and healthcare anti-inflammatory drugs (NSAIDs) in treating the signs 90 visits) in the United Kingdom and Germany. and symptoms associated with osteoarthritis (OA) and 80 adult-onset rheumatoid arthritis (RA). However, COX-2 • In-depth analyses were conducted to explore the cost 70 specific inhibitors are associated with a decreased risk of difference associated with ulcers and GI complications, 60 NBV-parametric gastrointestinal (GI) adverse events (AEs) compared with defined as GI serious adverse events (GI SAEs) in the 50 95% level nonspecific NSAIDs.1-3 clinical report. 40 5% level 30 • Pharmacoeconomic studies of COX-2 specific inhibitors • Cost-effectiveness ratios were calculated for cost per 20 have demonstrated economic advantages compared with averted gastroduodenal ulcer, cost per averted withdrawal

Probability cost-effective (%) Probability cost-effective 10 other treatments for OA and RA because of their favorable due to treatment failure and/or AE, cost per averted GI 0 4-6 SAE, and cost per avoided ulcer with GI SAE. GI safety profile. 0 2000 4000 6000 8000 10,000 12,000 • The COX-2 specific inhibitor valdecoxib is approved in • A symptomatic ulcer was defined as a GI AE reported as Value of ceiling ratio (€) Europe for the relief of the signs and symptoms of OA a gastroduodenal or peptic ulcer. and adult RA (10 and 20 mg qd) and for the treatment of • Total healthcare costs in the UK included the daily cost • Furthermore, the UK and German cost-effectiveness primary dysmenorrhea (40 mg qd). of valdecoxib at £0.77 and generic diclofenac at £0.16. planes show most of the valdecoxib samples in the right • In a previously reported 26-week, randomized • Total healthcare costs in Germany included the daily quadrants, indicating that valdecoxib is more effective in controlled trial (RCT) comparing valdecoxib 20 mg qd cost of valdecoxib at €1.3471 and generic diclofenac at avoiding AEs and sometimes less costly than diclofenac and diclofenac slow release (SR) 75 mg bid in adult €0.4659. (85% for the United Kingdom and 34% for Germany) patients with RA, valdecoxib demonstrated comparable • Bootstrapping was used to create a 95% confidence (Figures 7 and 8). efficacy, a lower withdrawal rate, and a superior GI safety interval (CI). profile to diclofenac (Figures 1-3).7 Figure 7. UK cost-effectiveness plane for symptomatic ulcer RESULTS avoided, valdecoxib 20 mg qd over 6 months.

Figure 1. Trial design. • The treatment groups were comparable with respect to 600 0% 15% age, race/ethnicity, and gender. Outcomes collected: 400 n = 246 Valdecoxib 20 mg qd • The clinical trial showed comparable efficacy for valdecoxib • Clinical measures – Efficacy and diclofenac, but a superior safety profile for valdecoxib, 200 (eg, arthritis pain) which resulted in fewer GI AEs and hospital days. – Clinical safety 0 Patients with (eg, ulcers) • Patients taking diclofenac were significantly more likely RA • Healthcare 200 resource use to be hospitalized due to SAEs during the 6-month study

– Hospital days Incremental cost (£) period (0.56 more hospital days per patient; 95% CI, -400 – Concomitant 0% medication 0.0006 to 1.1; data not shown). 85% n = 237 Diclofenac SR 75 mg bid – Study medication -600 – Unscheduled • Only 6% of the hospitalizations in patients taking -0.50-0.40-0.30-0.20-0.100 0.10 0.20 0.30 0.40 0.50 procedures valdecoxib 20 mg qd were due to GI SAEs vs 38% of the – Unscheduled 6-month treatment period healthcare visits hospitalizations in patients treated with diclofenac SR Incremental effect 75 mg bid (Figure 4).

Figure 4. Number of days spent at the hospital due to GI SAEs. Figure 8. German cost-effectiveness plane for symptomatic ulcer Figure 2. Clinical outcomes. 250 GI SAE hospital days avoided, valdecoxib 20 mg qd over 6 months. 95% CI of efficacy outcomes Non-GI SAE hospital days 200 1000 0% 66% -0.05 800 ACR-20 38% 0.02 150 600 Physicians’ global 0.1 (€) 400 Patients’ global

-0.81 100 st Tender/painful 200 joints score Hospital days 6%

al co 0 -0.9 Swollen joint counts 50 62% t 94% -200 Tender/painful joints emen 0.09 mHAQ 0 r -400 -0.62 1.74 Valdecoxib Diclofenac SR Inc -600 Arthritis pain (VAS) 20 mg qd 75 mg bid (n = 246) (n = 237) -800 Favorable for -6 -4 -2 0 2 4 6 Favorable for 0% 34% valdecoxib diclofenac -1000 -0.18 -0.13 -0.08 -0.030 0.02 0.07 0.12 0.17 0.22 Difference between valdecoxib 20 mg qd • The cost per averted gastroduodenal symptomatic ulcer and diclofenac SR 75 mg bid with valdecoxib in the United Kingdom was -£1104 (negative Incremental effect ACR-20, American College Rheumatology-20 Responder Index; number means cost saving for valdecoxib) and €386 in mHAQ-Modified Health Assessment Questionnaire; VAS, visual • The mean total cost per patient for healthcare in the analog scale. Germany (Table 1). United Kingdom was £452.07 for valdecoxib 20 mg qd Table 1. Incremental Cost-effectiveness Ratio Estimates over 6 months. and £556.63 for diclofenac SR 75 mg bid with a difference of -£104.57 (95% CI, -307.51 to 98.37). Figure 3. Safety outcomes. Analysis Outcome ICER in the ICER Variable United Kingdom (£) in Germany (€) • The mean total cost per patient for healthcare in Germany 95% CI of withdrawal rates and ulcer rates Avoided GD was €610 for valdecoxib 20 mg qd and €573.43 for symptomatic ulcer -1104 386 diclofenac SR 75 mg bid with a difference of €36.57 Avoided withdrawal due to (95% CI, -143.76 to 216.89). -7% Withdrawals treatment failure and/or AE -1580 553 Avoided GI SAEs -2709 947 CONCLUSIONS Avoided GI SAE -9% Gastroduodenal ulcers with symptomatic ulcer -3522 1436 • This economic evaluation suggests that the improved safety profile of valdecoxib is reflected in lower costs for GD, gastroduodenal; ICER, incremental cost per unit of benefit. healthcare utilization related to the treatment of GI Favorable for Favorable for valdecoxib -20% -15% -10% -5% 0% 5% diclofenac • The cost per averted withdrawal due to treatment failure complications. and/or AE with valdecoxib was -£1580 in the United • The differences in costs are mainly attributable to a Difference between valdecoxib 20 mg qd Kingdom and €553 in Germany (Table 1). decrease in hospital days and procedures related to GI and diclofenac SR 75 mg bid • The cost per averted GI SAE was -£2709 in the United SAEs for the valdecoxib 20 mg qd group vs the Kingdom and €947 in Germany and the cost per averted diclofenac SR 75 mg bid group. GI SAE with symptomatic ulcer was -£3522 in the United • Healthcare resource utilization data were prospectively • The differences are robust to changes in costing environment Kingdom and 1436 in Germany (Table 1). collected in this trial for the purpose of performing € showing the same trend in UK and Germany. economic evaluations to compare the costs of valdecoxib • The UK and German cost-effectiveness acceptability • The superior safety profile of valdecoxib compared with and diclofenac. curves for avoided symptomatic ulcer are presented in diclofenac translates into lower total healthcare costs Figures 5 and 6. for patients treated with valdecoxib and overall cost- OBJECTIVE effectiveness in both countries. • To compare the cost-effectiveness of valdecoxib 20 mg qd Figure 5. UK cost-effectiveness acceptability curve for symptomatic ulcer avoided, valdecoxib 20 mg qd over 6 months. and diclofenac SR 75 mg bid in the treatment of RA REFERENCES 100 based on prospectively collected data of healthcare 1. Goldstein JL, et al. Aliment Pharmacol Ther. 2003;18:125-32. 90 resource utilization in an RCT over 6 months. Cost- 2. Sikes DH, et al. Eur J Gastroenterol Hepatol. 2002;14:1101-11. 80 effectiveness evaluations were calculated for the United 3. Goldstein JL, et al. Aliment Pharmacol Ther. 2004;20:527-38. Kingdom from a National Health Service payer perspective 70 NBV-parametric 95% level 4. Svarver P, Aly A. Rheumatology (Oxford). 2000;39(suppl 2):43-50. and for Germany from a sickness funds payer perspective. 60 50 5% level 5. Peloso PM, Scheiman JM. Am J Med. 2001;110(suppl 3A):50S-4S. METHODS 40 6. Haglund U, Svarvar P. Rheumatology (Oxford). 2000;39(suppl 2):51-6. 30 • Clinical and healthcare resource outcomes were 7. Pavelka K, et al. Rheumatology. 2003;42:1207-15. 20 prospectively collected as part of a 26-week RCT of Probability cost-effective (%) Probability cost-effective 10 patients with RA receiving either valdecoxib 20 mg qd 0 (n = 246) or diclofenac SR 75 mg bid (n = 237). 0 200 400 600 800 1000 1200 Value of ceiling ratio (£)

NBV, net benefit value.

The study was sponsored by Pfizer Inc and Pharmacia Corporation. ISPOR 2004; Study 062 UK and Germany